What AI Still Can’t Do: Two Big Banking Narratives Get Tested

Hey, everybody, and welcome back to another episode of What’s Going On in Bryant Land, with your host, Stacey Bryant, and me, Ron Shevlin, just tagging along for the ride. That’s all I’m doing here, folks. We’ll get to banking in a little bit, but Stacey, how are you doing?

You know what, Ron? One day it’ll be Bryant Land. Until then, welcome to What’s Going On in Banking, y’all. I am doing great, despite feeling a little feverish and coldish. It’s September, so the kids are back in school. But you know what? I’m here for it. I’ve been traveling for a few weeks straight, but you’ve been traveling too. I want to know how you’re doing, my man.

Yeah, I’ve got the fever. No, what was that song? It’s too old for you.

Anyway, listen, the last time we recorded, you had just come back from Europe. You were in Spain. This time, we’re recording right after I got back from Europe—a week in Paris.

This is monumental for me, Stacey, because my wife is always on my case. She’s always like, “Hey, I have a great idea. Why don’t we take a vacation that’s not tied to one of your speaking engagements?” This was one of those. No work at all. It was great.

Here’s what I learned from a week in Paris: The travel books are all wrong. They’re wrong about where to go, where to eat and what to do.

Nancy and I averaged about 25,000 to 30,000 steps a day, which is a lot for us. But we got to see everything. Here’s my take. The places highest in the travel books, things like Notre-Dame and Sainte-Chapelle, were among the lowest of what I’d rank. There were so many churches cooler than Notre-Dame and Sainte-Chapelle.

Then there are the places they tell you to eat. Let’s be real here. Can I be honest for a moment, Stacey? Do I have your permission?

Okay. French food sucks. Can we just say that? It’s horrible. Escargot, all that stuff. Every French bistro has boeuf this and boeuf that. Listen, I love my boeuf, but it’s all the same.

We found some really good Italian restaurants to eat at. We found a little coffee shop with the greatest pancakes and fruit. It was amazing.

I decided I’m going to start a travel Instagram called Frick Steves. It’s the anti-Rick Steves. It’s, “Here’s where you really need to go, and here’s what you really need to see.”

I’ll give you a good example, and I won’t take long doing this. We were walking down the street, heading to some park we wanted to visit. We looked down a side street and saw this really cool church with gold domes on it.

We were like, “Hey, let’s go check it out.” It was late on a Friday afternoon. We walked in, and it was a Russian Orthodox church. They were in the middle—or really the beginning—of their service. There were no pews. Everybody just stood. It was so cool to see the service going on.

It was one of those things where, if you don’t take the detour and just follow the travel books, you’re standing on line at Notre-Dame, standing on line at Sainte-Chapelle or standing on line at the Louvre.

I figured, hey, I get to go to a number of places. I’m going to start my own travel blog.

There we go. We definitely want to hear that.

I can tell you a few things. I’ve been to France as well, specifically Paris. Like you said, when it comes to the churches, I was enamored by the architecture. I mean, Europe in and of itself, but especially the architecture of the churches.

There are so many different types of churches, and no church is similar to another. So I agree with you there.

I also don’t want the French to come for us.

Who cares? They’re not listening to this.

But let me tell you: Being a New York City girl, with that melting pot of all the cuisines, the best French food I had while I was in Paris was Thai food at this Thai restaurant, as well as an Italian restaurant that I think was Giorgio Armani’s restaurant or something. Everything else was liver-y. I couldn’t agree with you more.

I’m also very happy, similar to my trip to Spain during the summer, that you were completely shut off from work and got your steps in with your wife.

Are you ready to rock? Are you ready to rock and roll? Because I’m sure you’re recharged now.

No, I’m ready to keep traveling. I’m going to a user conference next week.

But hey, listen, you were on the road—well, kind of on the road. I guess you had to cross over the bridge from Jersey to New York. What were you doing?

First and foremost, I want to kick off with a big thank-you to Ty Muse, president and CEO of Visions Federal Credit Union in the New York–New Jersey area.

Ty is a good industry friend of mine, and for the past couple of years, he has been saying, “Stacey, you’ve got to come and be with our team for Paterson, New Jersey’s annual Dominican Day Parade.”

I saw him last week during Finovate week, which I’ll get to in a few minutes, and he said, “Hey, it’s happening this weekend. This is it.”

You know what I did, Ron? I took the girls and our Dominican flags, and we went to Paterson, New Jersey. We had to grab our pearls a little bit because, you know, we made it out of the hood. But I digress.

Nonetheless, it was amazing to see my people. It was amazing to see Visions Federal Credit Union come in deep. The team came in deep. We were all on a float. Sometimes I was on the float, and sometimes I got off the float and gave away Visions Federal Credit Union merchandise, bubbles and towels.

We walked what I think was maybe a two-mile parade. It was amazing.

You know I’m a community banking girl, and I always love to see a community financial institution walking the walk and talking the talk. They were there.

Shout-out to the Dominicans and to Visions Federal Credit Union for really embodying people helping people.

Then, as I mentioned, we had FinovateFall in Times Square in New York City last week. Another thank-you to Mark Meyer over at Filene and Tony Boutelle over at Origence. They had a FiLab fintech accelerator. Shanthi over at Casap was there. I met a few folks there, and it was nice to understand where these fintechs are really collaborating and partnering.

I think it was the next day when I had the opportunity and privilege to moderate a star-studded panel. Thank you to Katie Gwyn-Williams, Greg Palmer and the entire Finovate team, and to our guy Sam Kilmer, who was also in the building at the Marriott Marquis.

You remarked on this in a separate conversation: It was a super-long title for the panel. For the audience’s sake, I’ll read it out.

The name of the panel I moderated was, “Distribution Wars—Deposits, Embedded Finance and Platform Economics: How Can Banks Acquire, Retain and Monetize Customers in an Increasingly Fragmented Ecosystem? What Will Happen When AI Agents Gain Agency to Take Action on Behalf of Users and the Front End Is No Longer a Bank-Owned Channel?”

The four people onstage were Mary Miklethun, who runs business deposits and lending at U.S. Bank; Darius Wise, president and CEO of Red Rocks Credit Union; Patricia Montesi, CEO of Qolo, which, by the way, was just acquired by CSI, so kudos to her; and last, but certainly not least, TX Zhuo, general partner at Fika Ventures. He has backed so many fintech founders.

What I admired about him was that he made a whole bunch of avoidable mistakes, and he said he didn’t have a mentor while he was climbing the fintech ladder. He wanted to be what he didn’t have. Now he’s backing a whole bunch of seed-stage startups.

Here’s the thing: It sounds like a joke. I just need a priest in there. But you’d think putting a bank, a credit union, a payments builder and a VC onstage would produce four different flavors of panic.

Quite honestly, Ron, it didn’t. Nobody up there was rattled about the whole agentic future—AI working on behalf of customers. They had already made peace with the exact future the rest of the industry is still bracing for. That was a different energy that I really appreciated.

Every one of them talked about build versus buy versus partner. I want to tell you, it’s not a spreadsheet decision, whatever they tell you in the boardroom. We all agreed it’s a conviction decision.

Mary told me what it cost U.S. Bank to hand off its payroll experience to a third-party vendor called Gusto instead of building it themselves.

Patricia told me what it cost to turn down real revenue rather than compromise the architecture they had spent years building.

Darius over at Red Rocks decided that digital account opening wasn’t something they needed to own. It was something they needed to get right through somebody else, and he mentioned that they needed guidance there.

None of them sounded threatened by that math. They sounded like people who had already run the numbers and liked exactly where they landed.

But I want to hear your opinion on this. One thing really stuck with me, and it’s the basis of the conversations we’ve been having in every single episode. It’s the same idea playing out at a completely different scale.

I asked Mary over at U.S. Bank—she has been at the bank for maybe 27 years—what happens to her human team when AI agents gain enough agency to start acting on a customer’s behalf.

She didn’t give me some corporate “we’re leveraging synergies” answer. She told me about a janitorial company in Las Vegas, Ron. Check this out because it’s the whole ball game.

This janitorial company didn’t just clean buildings. Somebody noticed a very specific, unsexy, unscalable gap: cleaning up construction sites right before a property hits the market. They built a whole business around that one detail.

Mary’s point was that no AI agent optimizing for the fastest, cheapest and most efficient option is ever going to notice that. It isn’t sitting in the data. It’s in the conversation. It’s the curiosity. It’s a banker sitting across from a business owner long enough to hear the thing the business owner didn’t even know was worth mentioning.

Curiosity made money for this cat.

That immediately took me to new Cornerstone research on winning the business banking market, where eight in 10 of those businesses said they would seriously consider switching to an institution built specifically for their industry. Not cheaper. Not flashier. Just one that understands their niche.

I’ve been ranting for a while here, but Mary’s actual vision for U.S. Bank is almost insultingly simple once she says it out loud. AI doesn’t replace her team. It replaces the boring parts of the job—the paperwork and repetitive stuff nobody went into banking to do.

Her humans have more room to go find the next janitorial company. They have more time to be curious. They’re not less relevant. They’re more relevant because what made a good banker valuable was never the transaction. It was always the CQ—the curiosity, embracing ambiguity and having conversations.

Shout-out to all those panelists. That was my biggest takeaway.

Sounds cool. I love the CQ. Add that to IQ, EQ and all that stuff. It’s pretty cool.

I have a couple of reactions and thoughts on this. First, I’m a little confused by the title because it sounds like you were talking more about sourcing and supply than distribution.

When you’re talking about build, buy or partner, that’s input. That’s sourcing. Distribution is where you go with it. It seemed like a bit of a mismatch with the title, but I get it.

I love Mary’s point about identifying the opportunity because it supports something I wrote a couple of months ago about the creativity aspect of this.

I don’t know that AI agents and AI are going to replace all the boring parts. That feels a little too Pollyanna-ish for me. But I’ve been arguing—and I actually presented this at a conference in Quebec City a couple of weeks ago—that if you think about what an AI agent is programmed and designed to do, it’s a task-oriented thing. It is not a creative endeavor.

While generative AI tools help a little more with the creativity aspect, they’re formulating ideas based on existing inputs with no real context for whether something will work, how it would work or whether it has failed before. There’s no context that it brings.

It’s great for creating stuff out of the blue, but context is what humans bring.

I’ll share a quick story with you that has stayed with me for a long time. I always remember this from the days when I was a real consultant, unlike today, when I just work for a consulting firm.

I remember meeting with the senior management group of an organization. They were arguing about things and hashing everything out. A couple of members of my team and I were sitting there, observing as they argued.

One guy proposed a new way of doing something. Another guy said, “Well, we’ve tried that before, and it’s not going to work.”

The CEO jumped in and said, “Yeah, but this time you won’t be in charge.”

We have context, experience and perspective that we bring to a situation that AI just doesn’t have. It’s a manufactured idea. It cannot bring context and experience to the table because it’s formulating things from what it found and what the model was trained on.

I’ve never bought into the whole “AI is going to replace humans” aspect of this. None of that is proving true anyway. Employment numbers are actually kind of going up.

Yeah, we see Block and a couple of other technology companies say they’re laying off 30% or 40% of their staff because of AI, but they’re not. They’re laying off staff because they overhired. That’s basically what it is.

It sounds great, and I love the energy you guys brought.

I’ve got to tell you, Sam—whom you mentioned was also presenting at the conference—posted an amazing picture of you on the panel. You owned that stage, Bryant. I’m really proud of you.

You know, I learned from the best, Ron. I learned from the best.

We were in New York City. Listen, don’t put me on a New York City stage because, regardless of whom I’m presenting to, I’m going to make the best of it. It was amazing to be there once again.

I just want to mention this before we move to the next topic. It’s so funny because, again, I read you that very wordy, AI-generated-sounding title.

When I asked my panelists the specific questions in the title, all of them, in their own way, answered that the question isn’t about winning deposits at this point in Q3 2026. It’s about identifying and winning the overall workflow.

They all focused on that, which then led to, “Okay, if we’re going to own the workflow, there are components or steps within it that we can’t do on our own, or that it doesn’t make sense for us to build on our own.” That’s when it went to build, buy or partner.

I just wanted to say that. I also wanted to offer my own unsolicited opinion. It’s What’s Going On in Banking with Ron Shevlin and Stacey Bryant, so why not?

Finovate was a little underwhelming for me. I say that in reaction to something Sam Kilmer also posted and a lot of the conversations I was having in the hallways.

It was like, “Hey, my name is Joe Schmo. I’m at this particular company. Come visit us at our booth.” It just seemed so robotic, and I’m saying that with no pun intended. It sounded robotic.

There were a few folks who won Best of Show. But if I were to line up everybody who demoed, maybe 12—actually, 15—really solved, or clearly articulated how they solved, a pain point. Everyone else? I’m not sure.

I don’t know if AI is making it easier for folks to use some buzzwords and try to get into Finovate. But as we all travel to different types of conferences—whether they’re for banks, credit unions, fintechs or all three—and speak about technology, people invest their time and energy to come to these places.

Let’s really talk about how other people are doing it and what the outcomes have been so far. That’s what a lot of the community banking executives I’ve spoken with want. They don’t want pitches.

Tell me what you can do, what you can solve and what the outcome was.

Let’s stay on that point for a moment because I have to admit that it has been a few years since I attended FinovateFall, which I like a whole lot better than FinovateSpring. But there’s no need to go into that.

A couple of years ago, I kind of blasted Finovate—not the conference, but the attendees. The issue I had was that they were voting for the best user interface. They were voting for who had the snazziest interface, not who was solving business problems with real economics.

It was like, yeah, if your demo was slick and looked good, you got Best of Show—not because you were solving a business problem that had an economic impact.

What sounds positive to me is that the conversations you were having were with people who wanted that and not just the pretty stuff.

But here’s the challenge. If I’m a presenter at Finovate, let’s get real here. There is some currency to being named Best of Show at Finovate.

Can I tell you what I did? I went to Claude this week and said, “Go back about 10 years, look at the Best of Show winners and tell me whether there is an economic benefit to being Best of Show.”

It came back and said, “I cannot find a rationale supporting that.” The best it could prove was that there is certainly a PR and awareness benefit, but not a bottom-line benefit.

It’s not like the companies that have won can demonstrate that they went on to go public, get acquired or anything like that.

I had hoped it would find something because that would have made a really interesting story. But the reality is that there is some currency to being Best of Show. If winning Best of Show means having the prettiest interface, you’ve got a tough decision to make.

Still, it’s positive to hear that your conversations with attendees were more focused on economics.

Quick question before we move on to the next topic: Outside of the demos, were there any presentations that stuck with you or had an impact on you?

You know what? I’m super biased because I’ve seen her before, but I’ve got to give it up to Shanthi Shanmugam over at Casap.

Her background, as you know, Ron, is at Robinhood. She found a huge pain point around solving cases and disputes, including how financial institutions have certain deadlines. The way they’re leveraging AI blew me away, quite honestly.

Why? Because they are solving a major problem. That is why. That’s what blew me away.

Okay, so it’s Shanthi Shanmugam. I can’t pronounce it either, but I’m practicing.

The reason I’m practicing is that I’ll be at a Federal Reserve Bank of Boston conference next month, leading a panel on fraud in banking. Shanthi is one of my panelists, along with Ravi Loganathan from Sardine and Colton Pond from Socure.

I’ve got an all-star panel, and I’m really looking forward to that. But yes, Shanthi is amazing. That’s good to hear.

I’m watching the clock. I want to move into our next topic. Are you good with that?

Yeah, let’s do it.

Anything else on Finovate?

All right, good. We’ve done that one.

The big news this week—or maybe last week; I can’t remember the timing of anything, and by the time everybody listens to this, it’ll have been a couple of weeks—is that Chime acquired one of its banking partners, Stride.

I’d love to get your take. I’ll share my take on this, but let me preface it by saying that I’ve had a problem with Chime for 15 years now, or however long it has been.

I’ve always felt like its PR overstepped the boundaries. Back in the day, Chime was claiming millions of users or millions of accounts. It was probably true that when you became a Chime customer, you got both a checking account and an automatically opened savings account. Chime could easily double its account volume that way.

But there was a disconnect between its account volume and customer volume. I had been doing research on fintech adoption for years, and its numbers never jibed with mine. I kind of felt like maybe it was counting downloads. I don’t know what it was doing, but I’ve had an issue with Chime for years.

I’ve got to get over that. That’s my problem, not theirs.

But look at the logic and stated benefits in Chime’s press release. It was the AI story. It’s like, okay, there’s no AI story here.

Chime’s own legal firm, Wachtell, Lipton, Rosen & Katz—and actually, it was Ed Herlihy who wrote the brief—had an absolutely great brief talking about the logic behind the deal and what the benefits were. There wasn’t a single mention of AI in it. That stuff is all PR fluff.

Here’s what it comes down to: There are some real economic benefits that Chime gains by acquiring its partner bank. One benefit is that it doesn’t have to pay partner-bank fees anymore.

But here’s the issue I have with all this. Stacey, tell me where I’m really going wrong.

One of the reasons Chime’s CEO gives for the acquisition is that the partner-bank model has always been somewhat limiting from a fintech perspective because these are generally small banks. They’re all under $10 billion because they get the benefit of the interchange economics and can share revenue from those fees.

But they tend not to be technologically advanced or savvy banks, which has always been a limitation from a new product design and development perspective.

Here’s where I have a disconnect: I understand the economics of eliminating the partner-bank fees. It seems to me that all that money is just going to get plowed into renovating the tech stack of the underlying bank.

What am I missing?

I think I’m right around what you’re trying to feel out, and I took it a step further.

Before taking it a step further, shout-out to you for doing your due diligence, my friend. As you mentioned, you looked at the press release and said, “Wait, wait, wait. Hold on. Let me look at this legal memo.”

Listeners and audience, when you’re AI-ing anything and researching anything, make sure you look at everything. That’s exactly what you did, and it made me go down my own rabbit hole.

As I mentioned to you right before the call, I think it’s more money, more problems. You know how it goes: more money, more problems.

Here’s my prediction. I know you can’t stand predictions, but this is What’s Going On in Banking. I’m going to make a prediction, and I need you to challenge me because I’m also learning about this as we go.

I did a lot of my own research while trying to build this prediction, so let me try it and see whether it somewhat answers your question.

I believe this acquisition could slow Chime’s year-over-year growth and quietly weaken its valuation over time.

The moment you own a regulated national bank, you’re not a scrappy fintech with clever partnerships anymore. You’re a bank holding company that gets graded on GAAP profitability.

Last quarter, Chime posted $670 million in revenue against just $28 million in GAAP net income. That’s not a margin. That’s a rounding error.

I took it a step further. Here’s where Fed Daddy, 25 more basis points, walks back into the room.

Chime is leaning into lending. It’s already testing unsecured credit lines with higher-income members. As it inches toward that $10 billion ceiling, the pressure to grow lending revenue only gets louder.

But Chime is doing it right as borrowing has become more expensive for everyone. Higher rates squeeze the exact customers Chime built its entire brand around serving: people living closer to the edge of their paychecks.

So what happens here? I had to work through the banking ecosystem to understand this, but when you scale unsecured lending into a segment that’s already stretched, in a rate environment that just got tighter, what happens?

Putting my former credit union and banking hat on, I think what usually happens is that delinquencies climb quietly for a couple of quarters, and then it shows up as a write-down on somebody’s earnings call.

My overall read here, Ron, is that Chime built its fintech kingdom on being fast, cheap and regulation-adjacent. But now it has grown big enough that, just as the Fed is tightening the screws, the rules it’s about to inherit will collect their cut.

I think I might be right because the technology and compliance gaps at a bank like Stride don’t disappear the second Chime owns it. Stride is still involved.

I feel like there’s integration risk here, including absorbing Stride’s other fintech partnerships. The rate environment, as I mentioned several times, is working against Chime’s segment. There are a lot of ways for this to go sideways.

I was trying to challenge myself and see how I could be wrong, but I’d rather have you tell me where I could be wrong here.

I don’t think you’re wrong. I think you’re going down the same path I was, which is that there are just some things that don’t add up here.

Chime has benefited ridiculously from a valuation that I just can’t understand. Its customer base is so heavily composed of low- to middle-income consumers, which is why it hasn’t moved predominantly into lending. Look, that’s where the money and margin are in this business.

Number two, I’ve written about this so many times over the past five, six or seven years, encouraging Chime to expand its revenue sources into other types of noninterest income, fees and things like that. It has never done that the way other places have.

It’s just not adding up. Once it becomes Chime Bank, does it get valued as a bank? From a Wall Street perspective, what’s the basis for continuing to value it as a fintech?

I understand that there were fees involved in the partner-bank model, but there were two huge benefits. Well, maybe it’s the same benefit: You basically pushed regulatory compliance onto somebody else and kept it away from yourself. You didn’t have to deal with it. It was somebody else’s problem.

I get that there are fees involved. But becoming a bank did not solve Varo Bank’s challenges and problems. Varo has really struggled over the past couple of years.

None of this adds up for me.

By the way, regarding the whole $10 billion issue, a couple of folks pointed out to me on LinkedIn after I posted about this that when you start adding where Stride is, what Chime brings to the table and the other relationships Stride Bank has, they’re getting dangerously close to that $10 billion level.

The best thing that could happen to them is if somebody on the government or regulatory side of the equation got some brains in their head and realized, “Hey, we should raise the threshold from $10 billion to $25 billion.”

That would give everybody a lot more headroom. It seems to me that’s the best thing that could possibly happen to Chime here.

I’ve got to say, none of it is adding up. I don’t get it. It just doesn’t make sense.

More money, more problems. I’m sticking with more money, more problems.

I love it.

Okay, we’re really dragging this one out, so let’s put a stake in the ground, Bryant, with another episode of What’s Going On in Bryant Land.

Thank you, everybody, for listening. I hope you’ll join us for the next episode of What’s Going On in Banking.

Thanks, guys. If you enjoyed today’s episode of What’s Going On in Banking, make sure you hit that follow button and subscribe wherever you listen. Whether it’s Spotify, Apple Podcasts or YouTube, we’ve got more witty, gritty conversations coming your way, and you won’t want to miss them. Stay tuned.