Would You Let an Agent Be Your Muse? Or the Other Way Around?

Hey, everybody. Welcome back to another episode of What’s Going On in Banking? I’m here today with Stacey Bryant, co-host, to talk about what’s going on in banking. Stacey, what’s going on? Where have you been? We haven’t talked in a long time.

Yeah, I feel like it’s been forever, even though it hasn’t been. Nonetheless, I have been doing fantastic. This past weekend, basketball season started, Ron. I’ll be honest with you: My girls, eight and 10, started basketball, and they lost both games. They needed that. They needed to feel the competition, baby. They needed to understand and identify the heart of what’s out there. They needed to basically resurrect themselves.

I did catch myself going a little crazy as the coach-parent—not a coach, per se, but I thought I was one. I was directed to the parents’ lounge, where I saw a big sign that said: Number one, you are not the coach. Number two, this is a game. Number three, you and your kids do not play for the Knicks. That humbled me.

Outside of that, I’m ready to rock and roll. I’ll be at New Jersey Bankers’ Banking on the Future. By the time this airs, it’ll already be done, but I’ll be emceeing that later today. I’m looking forward to that.

Wait, before you go on, let’s go back to basketball here.

Okay.

I’m sorry that the girls lost, but to me, the real question is: Did they dominate the game for their team, and did they stomp the opponent?

They were aggressive. One piece of constructive feedback I gave my 10-year-old was, “You need to be more aggressive. You need to have your eyes on the ball at all times.” At one point, she got a little too aggressive and made someone cry. But I said, “Okay, you know what? Just remember that this is the game.” I had to remind myself of that, too.

To be continued. On top of that, we’ll be dribbling every single day and working on defense, Ron.

Okay. I want you to share this story with them. I’m sorry to go off on this. It’s not banking-related, but I saw a LinkedIn post today, Stacey, and I thought you’d be really interested in it. This guy was posting about something he was trying to call the Scalabrine Effect.

Here’s the story. Apparently, there was an NBA basketball player named Brian Scalabrine. I guess he retired around 2012. I don’t even know what team he played for. In the scheme of the NBA, he was not particularly good. He averaged three points a game and only played at the very end of the game throughout his career, until he finally retired around 2012.

Apparently, after he retired and was playing in local YMCA leagues, he posted on social media, “Hey, I scored 60 points last night,” to which social media just roasted him. It was like, “Yeah, of course you scored 60 points. You’re a former professional basketball player playing against my fat, ugly uncle who can’t dribble the ball.”

But a Boston radio station apparently said, “Hey, we should really test this out.” They created a contest where they invited people to apply to play one-on-one against Brian Scalabrine. This is a guy who just sucked in the NBA—three points a game, probably averaged two minutes a game. What do you think happened?

Uh-oh.

He crushed everybody. They played 11-point games, and he basically crushed everybody. I think he gave up five or six points.

The point of the post, which I found really interesting, was that the difference in quality between the average person and literally the worst player in the NBA is much greater than the difference between the worst player in the NBA and the best players, like LeBron James or Michael Jordan. You know what I mean? We just don’t think about the differences in quality.

I won’t go into this, but the post discussed how a researcher solicited videos of surgeons performing gastric bypasses. Without seeing who the doctors were—just their hands—the researcher asked a panel of other surgeons to rate the surgeries. Sure enough, there was this huge discrepancy in quality.

This has nothing to do with banking, but your basketball example reminded me of it. Sorry for the detour. Hopefully, there’s one person out there who’ll go, “Oh, the Scalabrine Effect. That’s pretty cool.”

Anyway, back to banking and back to travel. I’m coming off some travel myself. I was in New Orleans last week for a technology company’s user conference. I’m really excited about this because I spoke about why financial health and wellness efforts are falling short and what has to be done. We have a report coming out in a couple of weeks. I’ll put that on the agenda for a future episode.

But I want you to get back to what you were saying, because I know you were going to launch into something New York-related, and I want to hear it.

No, you’re all good. I was about to ask how you were doing on these NOLA roads. I definitely wanted to hear about that. I saw you looking all pretty in print on LinkedIn. The folks over at Bankjoy took some beautiful shots of you, so I definitely wanted to hear about that. Keep me posted.

One more thing I wanted to mention: This past weekend, I received a special invitation, but unfortunately, because of basketball, I could not make it. Founder Imani Ellis has been hosting an extraordinary New York City entrepreneurial gathering for the past eight years, Ron. It’s called CultureCon.

CultureCon New York City happened this past weekend in Brooklyn, New York. Whoop, whoop—my hometown. It’s funny because Chase Ink, for those of you who may not know, is Chase’s small business card line. They were the major sponsor of this event, which drew more than 12,000 people.

These 12,000 people, Ron, are content creators or folks who are itching to become content creators—entrepreneurs. I was looking through the agenda, and many of the breakout sessions were about building your brand, how to monetize, how to gain and where and how to bank, without mentioning Chase, so to speak. I found that pretty interesting.

They added a career lounge, so I immediately thought about the future of banking—or the future workforce, for that matter. They had creative business stages with a whole bunch of celebrities. John Legend, Aaron Pierre and Michael B. Jordan were there.

They had Braindates, which were peer-to-peer pitching sessions designed to move attendees beyond inspiration and into actual deal-making. I chatted with some of the folks who went, and they mentioned that they had heard of CultureCon through ChatGPT and Claude. They were able to participate in many of these Braindates because they had the backing of these large language models, which really shortened the gap and gave them the confidence to say, “I found some type of business. I found some type of content. Let’s do something about it.”

I wanted to mention that because, a few months ago, we discussed how Chase’s deposit share is increasing because of its branch expansion strategy. And really, the guts of that, Ron—you remarked on this in one of your Substacks—is that they’re doubling down on small businesses and wealth.

This is just another branch, of sorts, where they’re getting involved in the community. They’re taking a page out of community banks’ and credit unions’ playbooks and hosting these types of events. I just wanted to share that.

Hold on. This is really important stuff. At first, I thought, “Well, where’s the banking angle to this?” But, Stacey, the banking angle is a lot more than just Chase hosting something.

Do you know what I call content creators? Do you know the term I use?

What’s that?

Bizumers.

Oh, yeah.

Half businesses, half consumers. These creators don’t have LLCs. They’re not incorporated. They may not even think of what they do as a legitimate business. It’s an income source and something they do for personal fulfillment or perhaps to supplement their income.

I’ve been writing about this for a while. I don’t know if I’ve told you and Hannah this, but I’m writing a book called Bizumers. We’ve got funding for the editing, and it’s going to come out in Q1. It already has about 40,000 words, so I’m going to call it done.

But it’s going to be a call to arms and a call to action for financial institutions about this opportunity. There aren’t many banking products specifically for this group of people. They’re not traditional small businesses, but their needs are so much greater than those of the traditional or average consumer.

There’s also something coming up in the news. I can’t share it because I know it’s off the record, but something involving bizumers is going to hit in the next week or so. We’ll bring it up in a future episode, but I’m getting excited about the book.

That’s huge, Ron. Congratulations. I can’t wait to take part in the book signing for Bizumers. I think this will be another, perhaps, third part of The Innovator’s Dilemma—bizumer style.

Yeah. All right, some bank-related stuff. A couple of things happened, and I wanted to bring them up with you on this episode. At first, I thought maybe these were two separate things, but the more I think about them, the more they seem like two angles on a single trend.

A couple of things happened in the past week or so. I’m sure you’ve been following the news and all the press around the AI apocalypse—how it’s going to destroy society. I mean, gosh, it’s getting out of hand.

But that stuff has now come down to the banking world, and a couple of things happened in the past week that I think we should discuss.

One was a research note from an investment management firm whose chief economist warned about what he called an “agentic bank run.” The idea is that, as we turn over the management of our financial lives to AI agents, it will increase the possibility and probability of bank runs. Agents will see that other agents are pulling money out of a certain bank and accelerate that.

The other piece of news was Facebook announcing its AI agent called Muse, which would actually do all these things for us. Let’s talk about all of this—both Muse and the agentic bank run. What’s your take?

It took me to several places. Ron, I want to touch on Torsten Slok, that firm’s chief economist, and extract some of what he mentioned. To build on what you said, he wrote that if every household used AI agents to optimize the return on its cash balances, banks could lose a large share of the cheap deposits they rely on to make loans, which would be a problem for the entire financial system.

That took me to where we are in banking today. Again, we say this time and time again: We get the opportunity to be in boardrooms with both credit unions and banks, Ron. It made me ask, “What do we—as financial institutions—need to do now?”

Some folks have already adopted generative AI, and now they’re trying to understand and better organize their data so they can adopt some sort of agentic AI framework, regardless of the department.

I’m going to make sense of it all in a minute, but I know we—or you, rather—released an agentic AI report not too long ago. There’s a realization in there about the timeline. It talks about 2023 and 2024 as the generative AI phase: faster content drafting, quicker summarization and improved email personalization.

Now, how do we, as financial institutions, use that so we can gain a lot of these bizumers we started talking about?

That took me to a graph I sent to you and Hannah not too long ago. There’s a graph that we’re going to show, if you’re watching this, illustrating how many fintech companies are offering higher yields on deposits than the banking sector. I’m looking at it now. I see SoFi. I see LendingClub. I see Varo. I see Wealthfront. I also see AlumniFi.

I know AlumniFi. AlumniFi is a sidecar banking model from a credit union. Shout-out to April over in Michigan and MSU Federal Credit Union. You can articulate it better, but it’s a sidecar, or side-banking model of sorts, that really focuses on the university’s alumni and the products and services they need. I know April is huge when it comes to fintech collaboration.

What this really brings out, Ron, is this question: As financial institutions that have been around for decades upon decades—some for close to 100 years—what does their appetite for fintech collaboration look like now in Q4? I know we’ll touch on that in a few.

I also want to touch on one of the use cases that spills into Meta launching Muse. This research report discusses how agentic AI is reshaping marketing. One of the use cases it describes is intelligent customer journey orchestration. An agent monitors behavioral and transactional signals—for example, declining direct deposits, loans or CDs nearing maturity and digital banking usage patterns—and triggers personalized outreach at that moment.

I went down a rabbit hole. Do you know where this took me, Ron?

I know.

It took me to your boy Chris Nichols at SouthState. This was, I believe, in late 2024 or early 2025, when they created their own large language model, AutoGPT. They gave it a list of product descriptions, rates and performance metrics, and asked it to raise $2 million in deposits using an email campaign.

It figured everything out, and after a little more than three weeks, it raised $2.3 million in new deposits at a rate of 1.75% at the time. That’s a little different nowadays. But that came from a little more than 5,500 accounts.

I touched on a lot of things here, but if I can summarize everything I said, when we think about the bank or credit union of the future, we really need to start thinking about not only our tech stack, but our appetite for fintech collaboration and these large language models adding an agentic framework so we can win, thrive and remain relevant.

I know I spit out a whole bunch, but I’m certain you have remarks.

First of all, personally, there’s not a chance in hell that I’m going to give Facebook $20 a month for its stupid agent. There’s not a ghost of a chance of that happening.

But that raises another, broader question—not just about my personal use and behavior. You have to remember that Muse costs $20 a month. With ChatGPT, if you want to connect your accounts through Plaid, that’s $20 at a minimum, or you have to sign up for the $100 plan. Same thing with Muse.

So it begs the question, Stacey: Are people really that stupid that they’re going to spend $20 a month to have their money moved—to have their $200 moved from an account earning 1% to one earning 3%—when they’ll never even make their $20 back? Let’s do some simple math here.

That’s one aspect I think a lot of people are overlooking. It bugs the hell out of me when some economist from an investment management firm isn’t even looking at the mechanics of how this is going to happen.

All these folks who think agents are going to take over and manage everything aren’t looking at the cost. It’s $20 a month at a minimum. Most people bitch and moan about having to pay $5 a month.

Yeah, I’m going to push back on that. The reason I’m going to push back is that, especially now, we’re thinking about interest rates increasing, with Mr. Fed Daddy increasing them not too long ago. People are trying to—

Mr. Fed Daddy.

Mr. Fed Daddy. We’ll just keep it there.

But you know what? I have a whole bunch of subscriptions, Ron. I don’t use just one large language model. I know we’re talking about two different things here, but for the sake of information—the right information, or what I believe is the right information—and using those large language models as tools to get me where I need to be faster, I will pay for it.

I believe people will pay an extra 20 bucks a month. I’ll give you one scenario. Let’s say I just received an inheritance, or I’m about to receive one, and I need to understand where I can store this money—somewhere it’s going to be safe, number one, or what I like to believe is safe, and number two, somewhere it’s going to yield a higher interest rate.

I’m busy. I’m on the road. I don’t have time to go to my community banks or research all of this. For $20 a month, I’ll have this assistant figure out where I need to put this money. I think that scenario may work for some people.

Think about people who are paying brokerages to manage their money. I know we’re not comparing apples to apples here. I’m going through the product offering of it all, but they’re paying much higher percentages to manage their money.

This may even apply to the bizumers in your book who are no longer turning to their communities for financial advice. They’re turning to TikTok. They’re turning to whatever comes easily to them, with the hope of yielding higher profits. That’s my take.

I’m going to push back on your pushback, because let’s think about how this is actually going to work.

First of all, Bankrate has been around for more than 25 years, and even the laziest person can go to Bankrate.com and find out who’s offering the highest rates. But does the money move to those companies? No. It doesn’t. It never has.

We make a decision that says, “Hey, I’ve got $200 or $400 sitting in savings. I could get a better rate, but I don’t want the hassle of dealing with another company.” Not to mention the fact that sometimes I need that money in my savings account when I don’t have enough in my checking account, and I have to move money over.

Second, the money sitting in that savings account does not represent 100% of many people’s savings. They already have money sitting in investment accounts, brokerage accounts and high-yield savings accounts, making more money.

The other aspect these so-called experts aren’t looking at, Stacey, is the fact that, look, people might not like JPMorgan Chase and the big banks, but these are not dumb people. You’d better believe that, within no time, there will be pricing algorithms and mechanisms that impose a fee when people move money out or when the balance falls below a certain level.

Is the agent going to take that into account, or is it simply going to look for a higher rate somewhere?

I have one more thing I want to throw out there. Look, trust is still important. When Google, Facebook, Apple, Amazon and everybody and their mother are offering an agent, whose agent are you going to pick? Are you going to pick the one you trust the most?

If your most trusted agent is Muse from Facebook, we are not friends anymore. I’m just joking about that one.

But my point is that banks are going to offer their own agents as well. Are those agents going to make the right decisions on behalf of their customers or members and move the money out? How are they going to move the money out? Are they going to look at the pricing structure for moving the money out?

It’s much more complicated than simply finding a higher rate, and that’s what ticks me off. Some economist says, “There’s going to be a bank run because of agents.” That’s not how it works. As a guy in the financial services industry, you should know that instead of putting this crap out there.

Listen, I always take all of this back to what our folks can do. You mentioned trust. Trust is still one of the largest core values of community banking, Ron.

If I’m a bizumer and I’m typing into my large language model, “Hey, listen. This is the type of business I’m in. This is the type of accessibility I need. I live here. What’s the best bang for my buck?” I want to find a financial institution—or a fintech, for that matter—that understands my business, understands the way I bank and makes it fast.

The biggest takeaway I believe our listeners should carry with them is this: When you think about your tech stack and how quickly a prospective customer or member can join, what are the limitations, in this case, for business banking? Are you equipped?

I’m sorry to beat a dead horse here. I know we’re a little over time. But it even takes me to Nubank out of Brazil, which applied to enter the U.S. It has a department, Ron—and we’ve touched on this before—that focuses solely on the clicking experience. How do we take 10 clicks down to three clicks?

Why does this make sense when we think about Meta, Muse and agentic AI? If your financial institution can ingest these accounts while staying on top of trust and community, making sure everything is fast and mobile-responsive, and offering attractive products depending on the type of persona you’re chasing, then hopefully the AI agent can pick that up and direct the user to the desired organization.

I’m trying to map it out as I explain it. Hopefully, that makes sense.

It makes perfect sense. But you touched on something that I think could be the really important, core part of this.

It’s not the bank run or the movement of money. If the AI agent is really that good and can open the accounts, Stacey, think about what this does. Three clicks, 10 clicks or 100 clicks—it doesn’t matter. I’m not doing it anymore. I have an agent opening the account. I don’t care about the UX.

I’m not telling it to move the money every week or every month to get a higher interest rate. I’m just saying, “Take away the pain of answering all the bank or credit union’s stupid questions.” They’re not stupid—I’m sorry—but there’s an endless number of questions. If I’ve got something doing it for me, think about the convenience aspect of that.

Now, Nubank doesn’t need a department worried about that. This is the important part when you think about user experience: The user is no longer a human. It’s software, and software doesn’t care whether there are three steps or 103 steps.

I don’t think the AI agent cares about trust, does it? Does it care about the community or Miss Linda down the block, who knows my entire family and business?

But the agent is a representative—an advocate—of the user. So the question becomes: How do we communicate to the agent, “This is what’s important to me”?

Stacey, it’s been more than a year since I tried this, but I remember going on Claude and ChatGPT a little over a year ago, testing them out and asking, “Who has the best checking account out there?”

It came back and said, “I can’t answer that question. We need to know what the criteria are.”

At first, I was like, “Well, you tell me. You’re the smart one here. You tell me what the criteria are.”

It needed to know more about my preferences, but even the things it asked didn’t delve into trust and experience. How important is it to know somebody at a branch? How important is it to have a branch close by? It didn’t ask any of those questions.

How is that agent going to work if it’s limited to the information available online? If it’s just, “Hey, find the best Reddit or Yelp ratings,” then we’ve got a lot of problems. Or, like the kids say, we’re cooked.

I’ll say this: While you were speaking, I asked Claude, “Hey, I’m a content creator—an up-and-coming content creator. Who has the best business account for me?”

To your point, it wants to get a breakdown. What type of content is it? Is it Instagram? Is it TikTok? Is it YouTube?

For the sake of time, I can’t answer these prompts right now while we’re recording. But I am going to answer them and report back, just for giggles, to see where this takes me. To be continued.

All right, we’re coming up on 30 minutes. I think if we go into another topic, we’ll drag this out too long. We’ll put a pin in it for the next episode.

Stacey, I want to hear about your experience with Claude, ChatGPT and whatever other models you use to find the best account. But we can’t use the term “bizumer” yet. It’s not out there. I mean, it’s out there, but people won’t know what it means just yet.

Anyway, good conversation, everybody. Thanks a lot for listening. We hope to see you and hear from you on another episode of What’s Going On in Banking?

See you.

If you enjoyed today’s episode, please follow us on Spotify, YouTube, Apple—wherever. Please do it, because this is the eighth time I’ve had to do this thing. So let’s get it done, people. Follow us. Come on, man.