Fintech Takes x Chime presents Banking on Primacy Episode 4: The AI Episode

Hello, and welcome back to our podcast series, "Banking on Primacy," sponsored by our friends at Chime. This is the final episode in our series, so of course, we had to make it about AI. Alignment is a word that you hear a lot in AI, particularly among AI safety researchers who spend all their time thinking about the remote possibility that AI might kill us all. That's alignment with a capital A. However, I think the lower case A alignment is just as important, although admittedly it probably wouldn't make for his compelling of a science fiction movie. AI agents are being built for customers. How can we ensure that those agents act in customers' best interests? AI agents are being built to supplement, and in some cases replace the work done by humans. How can we ensure that those AI agents amplify the values of the companies that built them rather than undermine those values? These are tough questions, but my guess today, Ryan King, co-founder of Chime, was up for them. In our conversation, we discuss Ryan's views on AI as a technology wave and how it differs from anything else he's seen over his long career working in software engineering and product development. We talk about how AI is already starting to change Chime's internal processes, products, and and customer experiences. And most crucially, we explore the relationship between business model and AI alignment, which I think is going to become one of the most important topics in financial services over the next decade. It was a wonderful conversation to end our series on, and I think you're really going to love it. So, without further ado, here is episode four in our podcast series, "Banking on Primacy." Okay, Ryan, thanks so much for coming on. Thanks for having me, Alex. Appreciate it. Deleted to have you on. We are in the home stretch, and obviously we can't talk about primacy and consumer finance and banking in 2026 without talking about AI. So, that is going to be the focus of our conversation. I'm sure it is the focus of your work life on a day-to-day basis in a pretty profound way. Obviously, as I said in the intro, you're the technical co-founder at Chime. So, given your background in product and engineering and the amount of time you spent in Silicon Valley, I was curious just to start very broadly. This is a question I like to ask everyone who comes on to talk about AI. How would you compare AI to previous technology waves that have swept through Silicon Valley and swept through the financial services industry? Because from my vantage point, it seems fundamentally different both in its capabilities, but also in the effect that it's already having. But I'd love to get your perspective because you're much closer to it than I am. Yeah, thanks, Alex. Great to be here. Well, it's a great question for me. I've been in and around Silicon Valley, basically my whole life. So, I've had a front row seat to all of the prior waves of platform shifts and technology breakthroughs. I've really been a builder through the PC wave. I remember when I got my first PC in the 80s, obviously internet, mobile cloud, and each one at the time for me felt sort of enormous. And it was. But AI is really different. It feels really different from all prior waves to me. And as I reflect on that, every prior wave sort of changed what you could build and how fast you could build it and introduce new platforms to build on top of. And it gave us new step-changed tools. And so in that way, it's similar. But I think the way it's different and feels different. I actually think AI is different from every prior human technological advance since the dawn of our species, starting with the discovery of fire. Yeah. And I think that's because it's not just a step-change, it's a slope-change. And so, what I, it feels like the pace of our human technological progress will increase from here. And I don't know if that was quite the same way with the prior breakthroughs that I've lived through. That's fascinating. Yeah. Yeah. Yeah. Like the way I look at it, the closest analogy, obviously, I didn't live through the industrial revolution, but I'm not quite that old. But that's sort of like the closest analogy based on what I understand of it to what's happening right now with AI. So, you know, before the industrial revolution, humans had tools to help them perform physical work, like a hammer or the loom or whatever. But human labor was at the center of physical production. Everything we produced required human effort and human time. And when we went to sleep, then nothing was was made. And so, I think that that all changed in the industrial revolution, right? Didn't just give us better tools. It reorganized physical production around machines, factories, engines, you know, energy, all that sort of stuff. And so suddenly a factory is producing the goods 24/7 rather than human effort being at the center. And so, I think that's how I look at what's coming and what's happening now with AI and AI is poised to do the same thing, but for knowledge work rather than physical labor. And so, I think we're right at that inflection point right now where companies are reorganizing their cognitive production around virtual factories, all with human supervision, right? So, at Shine, we created our own software factory that's in the early days where sort of machines do all the work and humans just supervise. And I anticipate that every craft in knowledge work from finance to risk to legal to compliance to marketing, engineering, etc. We'll go through that same transformation in the next few years. So, if you think about it, just to sum all that up, I think like the analogy to the industrial revolution, I would say that like industrial factories mechanized physical production and AI will mechanize cognitive production. That's sort of how I think about it. It's a great way to describe it because you're right. I mean, it does feel I'm not old enough to have been there for the industrial revolution either, but you know, the Gen Z years listening might be surprised to hear that, but no, I'm not quite that old, but I do feel similarly where it's like it's hard to find a modern parallel that we've lived through. It's the same. And I guess just to double click on that for a second, I mean, obviously like software engineering, I feel like is the area where we've seen the most kind of breakthrough so far, right? And like it's pretty recent. I mean, it's what like last fall when this really all started to become evident, can you take us inside a little bit of just like your day to day over the last what three plus years kind of watching this go from sort of interesting learning about transformer architectures and kind of kicking the tires on this, but also maybe being a little bit like how do we actually leverage this to like cloud code and like some of the stuff that we can do now. I mean, I would imagine that even within this relatively compressed time frame, as it's only been like three four years, even within that there are some pretty clear kind of mile markers where things really shifted. Yeah, for sure. And if I reflect on just the last couple of years, yeah, I sort of look at them and maybe three different chapters, at least from my experience in time, you know, chapter one sort of started when chat, you, T sort of came onto the scene three or four years ago. You know, and everyone sort of stopped what they were doing myself included and say, okay, well, this is this is interesting. And maybe this can help us, you know, this is an interesting new tool that can help us in ways that we haven't, you know, been able to use before. And so that sort of first phase was all about like using LLMs and using these sorts of tools to, you know, help us a little bit in knowledge work. And then sort of like chapter two was when really you had, you know, quite good you know, coding assistance come onto the screen, cloud code, cursor, those sorts of things. And that really took it from an interesting tool that you might use on the side to a tool that became at the center of, you know, maybe a third or a little bit more of our employees or software engineers. And so all of a sudden, everyone's like kids and candy stores using using this tool, you know, and just don't like give you some sense of the statistics there like in that in that first phase, it's sort of culminated with like virtually every single chimer person who works at chime uses AI, at least weekly. And then at the sort of end of that second chapter, you know, something like 84 percent of our code was developed with AI at chime up from 29 percent, you know, just a few months ago. And so that you see that sort of like rapid, rapid adoption of AI tools and especially with AI coding where it's actually the work product is getting produced with AI is sort of in a primary way. And then we're just at the sort of dawn of this third chapter, which started maybe a few months ago, which is really what I was sort of making the analogy to in the, to the industrial revolution, which is rather than just with humans, being at the center of software production, but using super-powered tools, we're starting to think about and building, starting to build factories. So we've built our first virtual factory at Chime. We call it Archimedes. It's named after the ancient mathematician and engineer who is sort of like the father of this principle called Mechanical Advantage, which is basically saying that machines can be for some other players of humans. And so this is like-- Give me a lever and a firm place on which-- Exactly. And I shall move the earth, yeah. Exactly. Give me a lever long enough. I shall move the earth. Yes, exactly. You got it, right? So-- and that was, you know, 200, 300 BC sort of range. So that's the code name of our AI native software factory. And with that, we can-- and we're in the early days of that. But with that, we can, you know, describe an outcome we want, describe a feature we would like to exist in the product or a fix that we would like to happen. And the entire process sort of kicks off from there with, you know, humans in the loop at the right points to be exercising judgment, you know, making sure what it's doing is good and right and compliant and, you know, make sure we have all the right controls. But it's fully sort of autonomously orchestrated. So that's sort of where I see things going. And then of course, you say, OK, well, just producing software isn't the only requisite thing to creating value and actually helping with our mission. You actually have to, you know, get members to understand it, get the world to understand it. And so you think about things like go to market and all this sort of stuff. And so that's kind of where we're going next, which is, OK, how do we create-- how do we redesign all those parts of how we work around, you know, a factory mental model of like machines at the center and humans in the loop at the right place. So that's kind of the third chapter that we're-- and just at the beginning phases of now. Yeah, I mean, I'd like to double click into that because I think that, you know, there's obviously a great deal of uncertainty and I think anxiety that kind of permeates this entire discussion, right? And I go back to the Satrini report as an example of just like, someone writes kind of a science fiction paper that he wrote. And frankly, for everyone who sort of knew the part of the economy that the paper touched on, they read that part and they're like, that's actually not how it works at all, right? Like that's not how food delivery works. That's not how the card networks work. Like, that's not how any of this works. But even with the paper being highly speculative and not necessarily grounded in a deep understanding of how all these different industries work, everyone's share price went down overnight. And like, oh my god, like maybe there are no modes for any business anymore. And so there's this sort of like overarching anxiety. And I think what I observe, building on your point about like virtual factories and the ability to like completely reorganize your business is this sense, which I think is grounded in reality, that like the very foundations of all knowledge work can fundamentally change and be done differently based on these tools that we now have that are just gonna get better, right? The same in AI Land is, this is the worst that this thing has ever been. - The worst to ever be. - Right, right. It will never be worse than it is right now. And so, you know, thinking that through, I mean, I think in some ways software is, and sounds like Archimedes is like a virtual factory design for software development as you were saying, software is, I think a very obvious place for all of this to start, right? One is everyone building these AI labs, they're in Silicon Valley, they spend all their time with software engineers, like software engineering as a discipline is very well understood by the people who are close to this technology. And software development is also, I was talking about this with somebody else the other day, it's a very sort of like well structured and kind of linear process already, right? Like you have a development environment, you have a QA environment, you have production, like you have steps for moving from one to the other, you have gates that you have to go through, you have ways you can test. The software you're building is deterministic and predictable and you can design ways to test it and did the code work. Well, did it produce this outcome or not? Like it's very structured. And I think that has all lent itself to us making a great deal of progress in software development. And not just aided by AI as you were saying with the second wave, but actually done primarily by AI, which is the wave we're moving into. And I think what's interesting is that sort of suggests a direction for the rest of the knowledge economy and white collar work. But I also think getting there more broadly outside of just software, and maybe we can use product development as an example since I know that's part of your world as well. Like as you were saying, there's other things that go into understanding customer needs, analyzing what's happening in the market, analyzing competition, iterating on the design of a product, interviewing users, testing products, like there's a whole set of disciplines that if you were to step back at a 30,000 foot level, kind of look similar, maybe process wise to coding or software engineering or anything else. But when you get down to the details, probably your squishier in different ways. And it might be because product managers aren't used to thinking through tools, maybe the way that software engineers are. Maybe it's because there's some sort of like more kind of a nebulous, hard to define element to certain types of jobs. I certainly like to think that content creation isn't just autocomplete for a sufficiently large language model. I'm having that assumption challenged in my own world. So as this gets applied in this third wave, and especially beyond just software engineering, how are you guys thinking at time about adapting your business to this and like, what have you learned so far? Yeah, great question. You were sort of almost hinting that like, could this podcast have been completely created by an AI? I really kind of have good questions. It might have been when people are listening to this and might not be Alex and Ryan talking, just heads up. Yeah, we're maybe not that far off from that. But yeah, to your question, I think we're starting with, software development is just one piece of how do you create value in the world? And it just happens to be the tip of the spear in terms of where AI has the tools of materially changed what we can accomplish and we can reorganize our cognitive work around machines. And so where do we go next from there? And is it the same formula? And I think every craft and knowledge work probably has a different starting point in terms of you are saying how much structure, how much of the tasks are done in the digital world versus the physical world. You mentioned like interviewing customers and these sorts of things. So they'll all go through the same sort of transformation, but they're coming from different starting points. I'll give you maybe a small example at chime. You mentioned a couple times in your question, like interviewing users, talking to customers. We have a very strong user research function at chime, as you might imagine, given that we build for everyday Americans and most of us who work at tech companies and in Silicon Valley are a little bit different. Yeah. And so, and that's a very like, you're sometimes going into people's houses and seeing how they live and seeing how they deal with money. And so it's like, you think, OK, well, how is the NLM going to do that for you or how you can reorganize that knowledge work around a factory or a machine? But that's actually one of the areas where we've had, at least as much transformation and how would you work to what we see in software engineering surprisingly. I'll give you a couple examples of that. We've trained basically simulations of different segments of Americans that behave and act and think and talk and have the wants and needs and hopes and desires of different types of people. And so you can ask them questions. Well, you don't have to ask them questions. Your LM can ask them questions and get answers that are fairly representative of what would be like to talk to a real person. So simulated research, of course, you have to always ground yourself and making sure it's not hallucinating. Yeah, sure. Right. But I think we've started to see that already. And so when we think about where to go next, it's like, well, what are all the steps between identifying a need in the market and actually filling that need and using that, filling that need to create an outcome for our customers and our business. And that's where we're going to start going after next in terms of reorganizing knowledge work around factories and machines. And then when you get to the sort of, I'll call it more of the true sort of GNA functions at a company. I'm thinking finance, HR, people, teams, the sorts of things. We also see big opportunities there, but the shape and the trajectory of how those crafts transform will probably look a little bit different. And it's more about empowering the humans in those crafts to do more, rather than those things necessarily being looked at as a bottleneck to creating valuable software. And so I think it's like, go after all the pieces of the puzzle you need to create business outcomes and valuable software and things that make things that people want. And then sort of secondarily, you can look at turbocharging other aspects of a company. That's sort of my view. - Yeah, no, that makes sense. What do you think just as a follow up, have you found any areas where you think AI might not kind of disrupt something that humans are doing? 'Cause the thing I always think a lot about is, I find the doom and gloom about AI getting rid of all jobs to be kind of nonsense. But I do think we will have to kind of radically reorient where value lives. Have you seen any early examples and it could be in software engineering, it could be in product or user discovery or marketing or anywhere else? Like examples of someone using AI in a really like important, productive way, but still sort of providing something that the AI really can't. - Yeah, I mean, and if we just stick to knowledge work for a second, 'cause then there's the whole, 80% of the world's GDP is in physical goods and services. - Right, right, right. - I think they're like relatively safe from, totally, from LLMs at the moment of those some point you'll get into humanoid robots and stuff. But if we stick to knowledge work, I think, you know, and I tell my teams this at Chime, the things that are gonna be most important to, that AI cannot replace in any time in my crystal ball in the near future is taste, vision, judgment, and accountability. So taste is like, is this good, judgment? Is this correct? Is it right? Is it the right thing to do? Accountability, you can't hold a machine accountable, right? And so my analogy to, you know, the Industrial Revolution would be like, if you were making clothes by hand, you were sewing a sweater. And then the sewing machine gets invented. Well, and then a factory gets created to produce clothes at large scale 24/7. Okay, well, the job of the human doesn't, isn't to sew the clothes anymore, right? But it certainly is to figure out what clothes the world wants and how many of them to make sure they're being produced at high quality by the factory. And to be accountable, something goes wrong. If the, if the, you know, shirt is screwed up when you buy it, someone, you're not gonna hold the machine accountable to that. So I think it's the same analogy in knowledge work where we're gonna go through that same transformation. And I agree with you like, the doom or is them around, you know, everyone's gonna lose their jobs and all that kind of stuff, which is only sort of being reinforced by, you know, companies laying off people, knowledge workers in ways, I think, right? It's the reason it's tough, right? Which I think is like not great for the brand of AI. And, you know, and people's excitement for it, right? Yeah, yeah. There's a huge narrative problem, I think. Yeah, I think it is a little bit of a narrative problem that I think we are creating in Silicon Valley. But yeah, no, I think just the job's gonna change. And the GDP per capita is just gonna rise faster, which means we'll all have more fun. It'll live better lives in, you know, and do less of the mundane things in our work. You know, I think if you showed what we do for work to someone a human 10,000 years ago, they would think we're playing a game. I mean, look at us right now, right? Like, and I think it'll be the same, you know, when we look forward. Yeah, it's a great point. I mean, I think about that a lot where I'm like, work would be unrecognizable to someone, yeah, 50 years ago to say nothing that I had five thousand years ago, right? I mean, it's so completely different. And humans have an endless capacity to sort of reinvent themselves. And the world in which they operate. I wanna hone in on accountability because I think that is a great word to sort of flip it around to the consumer perspective on all of us, right? Obviously, we spend a lot of our time in the tech industry talking about how AI's gonna make us all rich or it's gonna change how what we're doing or it's gonna change our jobs and it's gonna do all of these things. We probably don't spend nearly as much time talking about what customers or members are gonna do with it. Yeah, or what they're already doing with it. And the theory, which I'd love to get your take on that I've had for a while with AI in financial services is among other things that will make it much easier for consumers to shop for financial products and to move between financial services providers in search of the best outcome. And to me, like the word I keep coming back to that's gonna be a huge challenge for many companies in financial services is inertia. Like if you built your business on inertia, if you built your business on, you know, people are busy and this particularly applies to kind of the everyday American segment of the market. Like people are busy, they have a lot of stuff going on in their lives, they don't have time or resources to just optimize every part of their finances. So as long as we're good enough and we don't piss them off too much, they're probably not gonna leave. And you know, this is why even though we've seen a definite rise and we've covered this in earlier episodes in this series, a definite rise in switching behavior and a lot more kind of soft switching or quiet quitting where people are opening up other accounts. It's still probably not as much as you might expect given the interest rates that most people get on their savings accounts or the rates that they pay on their loans even though rates have come down and they could refinance, like you don't see as much optimization of outcomes and movement exactly as you would expect, I feel like it's only a matter of time before AI kind of fundamentally changes that. What's your view on sort of the impact of AI on competition in financial services and like the impact on consumers who are gonna hopefully be more empowered? For most Americans, their primary bank account is their most important financial relationship. Traditional banks held that position and they took it for granted. Chime was built differently. Fee-free built to succeed when members do and now America's number one banking choice with roughly 10 million active members. Chime Prime takes that further. 5% cash back, savings rates up to nine times the national average, premium travel purse and no fees. See how at chimeprime.com. - Yeah, it's a great question and I love your framing about inertia. When I think about inertia and financial services that sort of two things that come top of mind are, number one, we've had e-commerce for 25 years and yet it's a relatively small percentage of commerce in the United States. It's less than a quarter. So people go to shop in person, right? It's like, well, isn't it? So it's like, okay, that's mostly inertia and behavior, right? Paper cash, paper cash is still 15 to 20% of retail commerce. - Checks, right? - People are still writing checks. - Yeah, my right for two checks. My right for two checks last week. And so renew our passports for our kids. You had to write a check, right? - Right, right. - Oh my God, I'm a fintech founder in my wife's writing checks. And so yeah, I think you're right that consumer behavior is, it takes time with that inertia, change in consumer behavior sort of takes place over decades. So with that framing, I think that's right. Having said that, I do think you're right that to the extent that consumers are able to use AI tools provided by anyone to optimize their finances for them to act on their behalf. I do think we'll see sort of like switching costs go down a little bit, but like you said, the inertia's still there. There's no reason for someone should have their money in a Wells Fargo savings account earning, no interest. But yeah, it might overcome a little bit of that inertia. And the way we think about it is, we have a financial assistant that we call Jade that we're rolling out to our members now. In a way that we think makes sense for the everyday Americans. So we've worked super hard to earn and keep the trust of millions of everyday Americans. And so the challenge is, how do you take sort of regular everyday people who you said are, they're busy, they're living their lives, they're not thinking about AI. In fact, they probably have like a little bit of a negative - Totally. - Do you have AI? - Totally. - Yeah, trust it. - It's destroying jobs and it's all that destroying jobs. - It's destroying jobs - before we get into sales. - Sounds bad, right? So how do you build something truly great that can actually help achieve financial outcomes for them? And how do you build on that foundation of trust? And so, that's how we think about it. Our challenge is how do we leverage the trust that we've built, create something great that actually helps people get ahead in every dimension of financial progress. And then sort of ease them into it over time, right? It's not gonna be, there are very few people that fit into our customer demographic that would push the full self-driving money button today. - Sure. - Even from Chai, even as much as they trust us. And so that's a big leap. - Yeah, it's a big leap. But I think we'll get people there over time, but that's sort of how I look at it. It's like our challenge is to build great things and sort of ease people into that behavior change. So that we're not sitting here 25 years from now. knowing with still people you know parking their money in ways that don't makes sense. Right. No, I think that makes sense. And I guess as a follow up to that, the other thing I've noticed, and I'll use like the bucket of personal financial management as kind of an example of this, that broad term, many different things that could fit into it. But one thing I've noticed is there seems to be kind of a push and pull happening right now between the, let's say like the sort of general purpose AI chatbots. This could be chat GPT. This could be perplexity, what have you, that are starting to really get involved or interested in financial services. Right. And probably seeing the same thing that we all see, which is that, look, this is just all ones and zeros. I sometimes use the example of physical fitness, you know, like financial services is very similar to physical fitness and that like, it's based on outcomes. There's a high level of trust. It's very, very high stakes. It's heavily regulated. But one thing that I think is different about financial services compared to healthcare is like in physical fitness and health, at the end of the day, you kind of have to change what you eat or how often you go to the gym to get results. You can get all the tooling and all this stuff around it as much as you want, but you actually have to get off the couch and do it for you. Well, right. Exactly. That's the problem. And I think in financial services, I think one of the things that's attractive to the open AI's or complexities of the world is they actually view it as like, we can lift the weights for you. Like we can do stuff for you to change outcomes and you don't necessarily have to do anything. So it's a very attractive opportunity. And you see perplexity pursuing partnerships with data aggregators to bring like financial data into their chat bot. You see open AI make an acquisition in the PFM space. So I think like one vision of how this might play out is these general purpose AI tools have different sort of flavors for different categories of jobs to help users with. And I think there are pros and cons to that. The flip side is what I think what you just described, which is the sort of AI embedded within a trusted environment. Right. So Jade, right. Yep. And we're seeing this across the industry where it's like, no, our view of this is, you know, AI is a very powerful tool. But the best way for consumers to sort of interact with it and get value out of it is within this environment rather than as a general purpose standalone tool. I mean, obviously you guys have a bias here. But like can you kind of talk of me through your thinking on that question of those two different models in which one makes more sense? For sure. I think that it's clear to me why the sort of open AI's and the perplexities of the world would want to enable that sort of stuff inside of, you know, their tools. You know, the way we look at it is it goes back to what I was talking about before around trust. And also what we're talking about about inertia is like, how are you going to get regular people to trust somebody else to lift the weights for them in your analogy? Yeah. Yeah. And that is going to be the hard part. The hard part is not going to be, you know, like we said, LLMs and Jade is the worst it's ever going to be, right? So it's going to get better and better and better. And so we'll open and so we'll chat you be T and so we'll perplexity. Right. So the question is, who are people going to trust to start to act on their behalf? And I think that trust is not something that changes or is built quickly, you know, especially in financial services. Probably the same is true in healthcare, right? People don't switch doctors, you know, every day. And so I think that the foundation of trust is the reason why we believe that we think the way to get people to the outcomes that we want to get them to of improving their financial situations is to build on trust and in environments and places where and with brands that they trust and use that as the way to help ease them into you know, automating parts of their financial life and so forth. But you know, we'll see how it plays out, but that's sort of our view. Yeah. I think that makes sense. And it ties to another word we haven't mentioned yet. You said accountability before which I think is incredibly important. The other word that I think is really important here is alignment, right? And this is obviously it's a big word in the world of like AI safety and AI research, but even in a more kind of mundane main street sort of sense, I think, you know, a consumer might not ever use the word alignment, because not everyone is AI-pilled the way we are, but they will say things that basically mean the same thing, which is how do I know I can trust going back to your point? How do I know I can trust this AI that I'm getting the best advice that possible, right? And like, you know, financial services, I mean, it's been fascinating to me to watch the evolution of like where people turn for financial advice and kind of what they're looking for, but like the reality is, you know, go back to PFM. Early PFM apps failed, I think partially because they screwed up the business model thing, right? Like, we are going to sell users and leads to other providers and we're going to monetize a free service that way. And when I look at, I'll just pick on OpenAI and chat GPT, they have 900 million free users, right? Most of them probably won't ever pay for the service and they don't have other ways to monetize them apart from ads. And so I do think business model is such an important component of alignment with like these new technologies or new experiences because at the end of the day, like, you know, at a time, you guys don't establish trust based on the number of branches you have or kind of like the like traditional signals in banking for trust. We had this conversation in an earlier episode. You guys establish trust based on our business model. Everything we do is aligned for best outcomes for you. And that's like, that's how we make money. That's how you make money. Like, there's alignment here. Is it your guys is thinking that like, that's going to carry through in your AI strategy and kind of what you do in this business as well? Yeah, absolutely. And you're right. I think you got to go back to the underlying sort of incentive structures. And then when we're talking about alignment in this context of financial services, we mean, yeah, is the company providing this service? Is there a business model aligned with the outcome that the consumer wants? In our case, our members, you know, and if the outcome is I want to improve my finances, right? What is the alignment against that? And how do you compare the alignment that an open AI or a complexity would have against that outcome versus, versus chime? And so for us, it's pretty cut and dry. You know, we run the primary account, the most important part of a person's financial life. We spend a decade or more earning that those trusted, you know, primary account relationships for more than 10 million people. Yeah, we don't charge fees. We don't penalize our customers based on how much money they have or don't have or what their spending obligations are. We've constructed the incentives and aligned our incentives so that we win when our members win. And I do think that, you know, that's different than other financial relationships out there. I think the big banks typically aren't incentivized to serve 75% of America. And it's not because they wouldn't want to. It's just because they haven't set up their business models. So it's for that to make sense. And so that's sort of the, I mean, that's the genesis of chime in the first list. That's why we started, right? So, you know, you look at, you know, brokerage account trading, stock trading accounts, they just want you to trade more, right? They don't actually really care if the price goes up or down. Red, prediction markets, maybe even more so. I mean, don't get me started, yeah. Yeah, yeah. So a lot of products out there have incentives that are like misaligned with the goals that their customers have there to provide. They can still provide valuable service. I'm not saying that they can't be valuable, but it does come down to incentives. And I think that that is sort of intrinsically linked with, with trust. And so, you know, that's the lens through which we look at the lens through. We look at AI is how does it help us further the mission and drilled on this aligned business model? How do we use AI to help our members make actual financial progress in ways that it would have been hard for us to without it? So, like that's really a lens through which we look at it at chime. Well, let's end by talking a little bit about that because I will say I am a little disappointed in my industry broadly in the sense that the most exciting thing to people working in financial services, if you were to just do a survey of kind of all the headlines happening right now, seems to be agentic commerce, which is fine. And I've used LLMs to do a little shopping and a little research before. I think many, many people have. I can certainly understand why anyone who's in the business of facilitating payments would like agentic commerce? Like, gosh, if we could just, if we could remove the human who has this annoying habit of like putting something in the hand, but then like leaving or going to think about it, like if we could just take them a little more out of the loop. It would be better for all all of us, there'd be more money to influence with the pipes. - Right, right, I can certainly understand that. But I mean, one thing that's come up again and again in this series is you guys have a very thoughtful approach of not swinging at every pitch, not chasing every sort of hot thing that everyone else in Fintech is doing, but being very thoughtful about, okay, what is our mission? What are our members actually want and need? And then how does whatever this new thing is that we're looking at, how does it fit into that and like genuinely add value? So give me some use cases, and these can be things you guys are already doing with AI or things that you're kind of noodling on, but like what are some use cases for AI, for how members might benefit from it that might be a little different than the, what you normally hear looking at all the press releases? - Yeah, no, you're right. There's certainly a lot of excitement around agentech commerce and every major player, VSA Google is sort of racing to own that infrastructure layer. As we just were talking about incentives, we can all understand and imagine what their incentives are for wanting that to become reality. So, but most of that conversation is about removing friction from spending or completing transactions like you were saying. And for me, the more interesting question is sort of what you were hinting at, which is like, is making it easier to spend money? Really the number one problem that Americans face? - I doubt that's what your members are telling you and what I hear. - That's what I hear. - Right? Yeah, that's not what I hear. - Yeah. - No. That's not what I hear. And then to add this to what I said earlier, online commerce is less than a quarter of retail spend, right? So even if making it easier to spend money were the biggest problem, you probably wanna figure out how to make it easier to spend in the real world first, right? Send a humanoid robot to Walmart for you. I don't know what that looks like, but we're not working on that. - I'm sure there were, I'm sure someone else is working on that. Yeah, sure. But yeah, when we think about this at time, it's less about making it easier to spend and more about spending smarter. So, you know, that's why that's what we're designing, J2B, it's designed to help you make financial progress. So whatever that might look like in your specific situation, maybe the most important thing is to start an emergency savings fund. Maybe you have some opportunity to consolidate high interest debt. Maybe you're putting money into an investing account, but it would actually be better if you use that to pay down interest debt. - Sure. - You know, and so when it comes to spending, the way we think about it is not making it easier to spend, but about saving you money, making you spend smarter, not making you spend more or more easily. You know, are you taking advantage of all the deals and discounts and are you maximizing your rewards? And, you know, maybe you can get a few bucks off at subway, but you need to realize that when you just, you know, because you're busy, you're living your life right. And that's where we see the opportunity for AI and for J to sort of help facilitate all that stuff and actually help you spend less and smarter, not spend more and easier. And so that's our our vision for J, it won't just be a chat bar or, you know, a PFM to something you ask questions about what your budget is. It'll actually be the proactive assistant that automates some of these things and your financial progress. And it'll be the lens through which you look at, okay, I'm busy, I don't have time to optimize all these things. Can you help me do that? And I think we're a well position for that, 'cause the sort of primary account being the most important financial relationship, we think this creates opportunity for us to build something entirely new that actually helps people that banking and Fintech world sort of has hasn't seen before. - I love that. I mean, and I think even like the ability for that discussion that a member can have to be like hyper personalized to them and their situation to be something that can flex to be as sort of like therapeutic or not as it needs to be. Some people want like very cut and dried answers. Some people want to sort of like have a bit more of a conversation and we just, we haven't had capabilities that provide that level of sort of flexible intelligence that we now have. So it seems to me, and I guess you guys are thinking about the same way that there's just a tremendous amount of opportunity there. And you know, the good news for you guys is if everyone's working on agentech commerce and making it easier to spend money, not as much competition on the other side. So that's an opportunity. - Yes, yes it is. And you know, yeah, sometimes people need a coach that gives them tough love and sometimes they need a therapist and you know, we'll build both of those and everything in between to help people. - I love it. Well, that is a great place to leave it. Ryan King, thank you so much for coming on the podcast. This was great. It's been a pleasure, Alex. Thanks so much. I enjoyed the conversation. - Thanks for listening. Today's episode was brought to you by Chin. JD Power found that more Americans open new checking accounts at Chin than at any bank or FinTech in America, two quarters in a row, 97% of Chin members say it's unlocked their financial progress. Chin Prime now builds on that. Expect more cash back, higher savings rates, travel perks, and still no fees. Let your hard work pay off at Chin Prime.com.