Hunger Games Banking: Nubank’s U.S. Play, AI Work Slop, and the Core Wars

Hey everybody, and welcome back to another edition of the What’s Going On in Banking podcast. I’m Ron Shevlin, chief research officer at Cornerstone Advisors, and here, of course, with my co-host and Cornerstone’s director of client development, Stacey Bryant.

Stacey, how are you doing today?

I’m great. I’m finally feeling healthier. I think I told you last week that I caught a bug. It’s fall here in the Northeast, but I’m happy to be bug-free and back at it.

I’ve already been on the road. Earlier this week I was at the New York Bankers Association, where they wanted someone to talk about AI, the omnichannel experience, and how to keep the human element at the center of community banking.

I’m continuing to hear the same questions from community banks: What does AI actually look like for us? How are other institutions using it? Where do we start?

I also had a similar conversation with smaller credit unions in the Maryland and D.C. area. So that’s been my week after recovering.

How are you doing? What’s going on with you?

I can’t complain, other than the fact that I travel constantly these days. October looks like it’s going to be a brutal month for travel, but that’s the way it is.

I just got back from a technology company’s user conference. It was a relatively small group, but the attendees were very senior, including CEOs, CFOs, and chief operating officers, and the main topic was data.

It was great to be in a room full of fellow data nerds and data geeks.

We actually did some research for them that should be published around the time this episode comes out. There were some interesting findings, and we’ll get into those later.

But first, let’s talk about something more recent that has become a meme in the social-media world: workslop.

I don’t know if you’ve heard the term before. It came out of a Stanford-related study and refers to AI-generated work that looks polished but lacks real substance.

The researchers argued that this kind of content is creating productivity losses and estimated the cost at about $186 per employee per month. They also found that people who produce workslop are viewed as less creative, less capable, and less reliable.

I doubt it will surprise you that I have some problems with the study.

On the positive side, I agree with the observation that generative AI lowers the barrier to creating content that looks passable but is shallow.

These tools can produce fluent, grammatically correct writing that still lacks analytical rigor, originality, or accuracy.

And I give the researchers credit for trying to assign a dollar amount to the problem. Most studies make broad claims about productivity and stop there.

But I think they may be over-attributing declines in work quality to AI.

How much has the quality of work already changed over the last 5, 10, 15, or 20 years because of education, social changes, staffing reductions, workload pressure, and other factors?

If an organization cuts headcount and expects fewer people to do more work, that can degrade quality regardless of whether AI is involved.

I also think the definition of workslop is tricky.

I use ChatGPT, Claude, and Perplexity all the time to generate first drafts, pressure-test ideas, and help brainstorm.

Where exactly do you draw the line between a rough draft created with AI and “hollow” content?

So I have some issues with the study. What do you think about the whole workslop topic?

The title is definitely catchy.

And yes, I’ve received workslop myself, especially from people reaching out to me on LinkedIn.

I’ll get a message that starts with something like, “Stacey, I love all the work you do for Cornerstone Advisors,” and then immediately dives into a sales pitch that clearly wasn’t written for me.

It doesn’t compel me. I just keep moving.

Like you, I use ChatGPT, Perplexity, Claude, and other tools. But I use them with context.

I’ll say, “Here’s the draft. Here’s what I’m trying to communicate. Here’s the audience. Here’s the tone I want.”

And I can be a little snarky, so depending on the recipient, I may even tell it to reduce the snark.

That’s how I think these tools should be used, as an editor, thought partner, or proofreader, not as a replacement for the reason you were hired in the first place.

I was reading the comments on your LinkedIn post about workslop, and Doug Bohenhouse from Basefound AI made a point I liked: context is everything.

People have expertise, networks, institutional knowledge, and experience that took years to develop. You and I each have decades in financial services. We know what has worked, what hasn’t, how the processes work, and who the people are.

The value comes from using AI to amplify that expertise and creativity, not to replace it.

So that’s my takeaway. Don’t ask the tool to simply “do the work.” Give it the context that makes the work valuable.

A couple of thoughts.

From a management perspective, I think leaders are going to have to make some uncomfortable calls.

If somebody submits something and you know that person well, sometimes you can immediately tell they didn’t really write it.

LinkedIn drives me nuts because you see these posts that are obviously AI-generated, not even because of the content, but because of the structure. Human beings do not naturally write in endless little punchy lines with perfectly spaced emoji bullets.

As a manager, you may have to sit someone down and say, “I know this isn’t your work, and this doesn’t fly.”

That creates more than a reputational problem. It creates a ripple effect inside the organization because people begin using AI as a form of intellectual outsourcing rather than as a productivity tool.

The way I use these tools is different.

I’ll go to ChatGPT and say, “Here’s what I think. Tell me why I’m wrong.”

It will push back, and sometimes I’ll realize there’s an angle I hadn’t considered. Then I may incorporate that by saying, “Here’s the other side of the argument.”

That thought may have been stimulated by AI, but I’m still making the judgment about whether it belongs in the work.

I don’t think anybody has accused me of workslop yet, but I absolutely use these tools.

So there is a productive way to use them. Managers just have to stay attuned to the difference.

It makes me wonder what ghostwriters did before ChatGPT. What happens to them now?

But let’s move on.

A few episodes ago, I talked about Nubank, the Brazilian digital bank that has become the largest digital bank in Latin America.

Well, according to American Banker, Nubank has now applied for a U.S. national bank charter with the OCC.

That fits into its broader global-expansion strategy.

The company’s stated objective is to serve both U.S.-based customers and people with ties to Latin America, including expatriates.

The charter application would cover a full suite of products: checking, deposits, credit cards, lending, and digital-asset custody.

Nubank already serves more than 120 million customers across Brazil, Mexico, and Colombia.

Its pitch is built around a cloud-native, low-cost, customer-centric platform. These are the same people who obsess over reducing clicks and making the mobile experience simpler.

The company’s leadership also has deep experience with the development of Pix, Brazil’s instant-payment system, which we discussed in an earlier episode.

That original design work focused on concepts users cared about: cheap, fast, transparent, open, and secure.

Nubank’s founder and CEO, David Vélez, has said the company remains focused on growth in Latin America, but the U.S. charter creates a path to serve people here and eventually attract new customers who value a digital-first offering.

Co-founder Cristina Junqueira has relocated to the U.S. to help lead the effort.

If approved, this would let Nubank offer products directly without relying on partner banks.

So I have two words for this: Hunger Games.

The financial-services Hunger Games.

What do you think?

I have a few reactions.

First, Nubank is hardly the first foreign fintech or digital bank to try entering the U.S.

Over the last decade, several European companies came in and largely failed to gain meaningful traction.

I remember interviewing the U.S. country manager for N26 when it entered the market. I asked, “What is the big problem in U.S. banking that you’re going to fix?”

The answer was essentially, “The mobile-banking experience in the U.S. is terrible, and ours is much better.”

I remember thinking, “You’re delusional.”

The digital-banking experience in the U.S. is actually very strong at many large institutions. Chase, Wells Fargo, Bank of America, and others have sophisticated mobile platforms.

And consumer perception matters. Even if the experience is better somewhere else in the world, most Americans have never used those products, so that comparison does not automatically motivate switching.

What is different this time is scale.

Nubank and Revolut are not tiny startups trying to enter the market. For all practical purposes, they are larger than the vast majority of U.S. banks in terms of customer scale, resources, and technology.

That gives them the ability to absorb the cost of a charter and build directly.

Varo was one of the first U.S. fintechs to pursue a bank charter, and you could argue that the cost and effort hurt the company. It may have been better off staying with partner banks longer while building a larger customer base.

Nubank and Revolut can afford to make that investment.

The other thing that is different is digital assets.

I don’t care whether Nubank has 10 million or 100 million customers in Brazil. That alone doesn’t guarantee anything in the U.S.

But if it enters with differentiated capabilities around digital assets, tokenization, or crypto, that is a more open field in the U.S. and could give it an opportunity to leapfrog some incumbents.

There was also a good LinkedIn post from Kabir Kumar at Flourish Ventures about Nubank’s charter application and the broader shift in U.S. policy.

He argued that simply submitting the application shows how far the U.S. has come in enabling financial innovation. He also said the best consumer policy is one that creates room for innovators to deliver.

I agree with some of that, but I’d push back in two ways.

First, I think we’ve come this far largely because the political and regulatory environment changed. If the 2024 election had gone differently, I’m not sure we would be seeing the same wave of charter applications.

Second, a lot of the financial innovation entering the U.S. is actually coming from outside the U.S., from companies such as Nubank and Revolut.

So yes, there’s more room for innovation, but some of the most aggressive innovators are foreign companies entering the market.

That’s fair.

What interests me is the niche.

Nubank has a natural connection to Latin America and could use that to serve people in the U.S. who have financial ties across borders.

There are millions of people living outside their country of origin, and that creates very specific financial needs around payments, banking, remittances, and digital assets.

That’s a real segment.

And beyond expatriates, younger consumers increasingly expect digital freedom. They hate friction. They may want their crypto wallet and their mortgage in the same ecosystem.

So the question may be less, “Will Nubank steal everybody’s deposits?” and more, “Why would a customer change financial providers while watching a fintech story unfold on TikTok?”

For community institutions, this should be a reminder to modernize.

If your brand is still built primarily on free coffee and donuts in the lobby, you probably need another differentiator.

Wait, there are banks offering free coffee and donuts? How have I missed this?

My wife probably knows I’d just go in, eat, and leave.

So what I hear you saying is that this is not only the Hunger Games. It’s the younger Hunger Games.

That was terrible. I shouldn’t have said that.

All right, ready for the third topic?

Let’s go.

The last couple of weeks have seen an explosion of acquisitions among core and banking-technology providers.

FIS acquired Amount. Jack Henry announced an acquisition. Fiserv announced two deals, including StoneCastle and Smith Consulting.

Let’s start with FIS and Amount because I know you wanted to talk about that one.

The way I look at FIS acquiring Amount is as a strategic move to make its banking software more flexible, cloud-native, and digitally competitive.

Amount has processed more than 150 million applications across deposits, loans, and cards.

By bringing that technology inside FIS, FIS can offer more unified digital-origination tools to its financial-institution clients.

That can help banks compete for deposits, grow lending and card portfolios, and modernize their digital experiences.

So from my perspective, this is a smart acquisition. FIS understands where its clients need to improve and is doubling down on it.

Agreed.

Now let’s talk about Fiserv and StoneCastle.

On the surface, you could look at it as Fiserv buying a deposit network to compete with providers such as IntraFi and Ampersand.

Fiserv’s chief operating officer said the acquisition strengthens Fiserv’s core-banking capabilities and gives institutions more options for cash custody, including supporting Fiserv’s stablecoin strategy.

But when I read that, I couldn’t help wondering whether the acquisition may eventually be even more valuable to Fiserv’s merchant clients than to its financial-institution clients.

You can imagine Fiserv using StoneCastle to build a stablecoin sweep service or treasury hub for merchants and retailers.

I also wondered whether there was a Finxact angle. StoneCastle already uses Finxact for parts of its digital-asset business, so maybe Fiserv sees this as another way to accelerate digital-core use cases.

Those are speculative thoughts, but overall I think it’s a strong move.

The way I initially looked at it was as part of the war for deposits.

StoneCastle is essentially a network that helps institutions place and manage deposits safely across multiple banks.

So I saw Fiserv buying it as a way to give its bank and credit-union clients easier access to that kind of deposit-management infrastructure.

I see that, but I’m thinking more about deposit management than deposit acquisition.

Maybe StoneCastle is simply a good business and Fiserv likes the economics.

But I also think it could support Fiserv’s stablecoin and treasury ambitions, particularly on the merchant side.

Both FIS and Fiserv have spent years expanding beyond traditional core banking and deeper into merchant services.

That’s why I think the merchant cash-management opportunity may be the bigger long-term play.

And this is consistent with what we’ve been writing about in What’s Going On in Banking for years. The biggest providers keep acquiring capabilities that let them extend further across the financial-services value chain.

There’s also a broader pattern here.

Historically, the big core providers waited until a technology category matured, then acquired one of the established winners.

Now the pace of change is forcing them to make bets earlier on emerging technologies and providers.

Bankers may not love that consolidation, but there can be a benefit if the acquisition results in easier integration with core and digital platforms, more capital behind the product, and greater long-term stability.

Banks are understandably cautious about relying on tiny technology companies for mission-critical functions.

So there are tradeoffs.

But the broad direction is clear: the big are getting bigger.

All right. I think that’s a wrap for this episode.

Stacey, great talking with you.

Thanks to everybody for joining us. We hope you’ll listen to another episode of What’s Going On in Banking.

If you enjoyed today’s episode, make sure to follow us on Spotify, Apple Podcasts, YouTube, or wherever you’re listening. We’ve got more conversations coming your way, and you won’t want to miss them.

Thanks for tuning in.