80% of bankers are concerned that AI can lead to misinformation and biased decision- making. But sitting on the AI sidelines means missing out on significant operational and competitive opportunities. How can bankers thread the needle? Nima Ghamsari, Co-founder and Head of Blend, joins American Banker Senior Industry Analyst Michael Moeser to discuss how financial institutions can avoid AI pitfalls while taking advantage of near-term benefits. Ghamsari will share how AI could transform banking within the next five years, and how financial institutions can bridge the gap between today's immediate opportunities and what's possible in the future.
Transcription:
Transcripts are generated using a combination of speech recognition software and human transcribers, and may contain errors. Please check the corresponding audio for the authoritative record.
Michael Moeser (00:09):
Despite all the hype and media attention about generative AI, since it burst into the mainstream spotlight almost two years ago, consumer and business sentiment indicate that we're still in early days of the technology's usage and adoption. Recent studies by American Banker reveal that banks are just starting to investigate and educate themselves about the breadth of generative AI based solutions that are becoming available and how they can best be applied to support their businesses. In fact, only a limited number of institutions have taken the next step to test or pilot these new technologies. Even fewer organizations have taken the bigger step of rolling out generative AI into their organizations, among employees and customer facing touchpoints. For all the hope and possibilities that come with generative AI based tools, there's a high degree of mystery and distrust about the unknowns, which may be holding people back from further exploration.
(01:09):
A case in point, roughly 80% of bankers are concerned that AI can lead to misinformation and biased decision making. But does sitting on the AI sidelines mean companies are missing out on potentially significant operational and competitive opportunities? Well, that speaks to the heart of today's discussion. Welcome to AI and Banking Science Project, or long-term growth strategy hosted by American Banker in partnership with Blend. I'm Michael Moeser, senior analyst at American Banker and your host. Today I'm speaking with Nima Ghamsari, Co-Founder and Head of Blend. For background, Nima holds a US patent, took blend to an IPO in 2021 where he actually rang the opening bell at the New York Stock Exchange, a former Palantir engineer and an alumnus from Stanford University. Welcome, Nima.
Nima Ghamsari (02:03):
Thanks for having me.
Michael Moeser (02:05):
Awesome. Well, okay, Nima, I understand you used to be a professional poker player. So this first question is right up your alley, similar to when you would walk into a room and you'd have to size up the players and make a decision of what are the opportunities, what are the challenges? Do I sit down and play or do I watch from the sidelines? What do you see as you gaze upon the generative AI buffet as to what's there in terms of do you sit down and play or stay on the sidelines?
Nima Ghamsari (02:34):
Yeah, for what it's worth, I was a professional poker player in 2004 to 2011, so it's been a while, but even that game is being transformed by AI. You can now run simulations and AI can help solve hands that previously could only be solved by humans. And so watching that space evolve and people acknowledge and understand and appreciate that, it sort of, I think is a precursor to what I think we're going to see in banking where there is so much opportunity. AI is evolving so rapidly that it can do things today that it couldn't do a year ago, and in a year from now, it'll do things that it can't do today. So I guess my baseline view of this is it's a space that we have to pay attention to and we not only have to pay attention to, but there are areas of, if you're running a bank or you're running a credit union or whatever it is, whatever part of the business you're in, figuring out the areas that you think are opportunities that maybe you don't know if AI could solve it yet, but areas that you're not as efficient as you could be.
(03:41):
You're not using as much technology as you could be. Those are the areas that I think AI will slowly creep into over time, and it won't be a short-term investment. It'll be a long-term investment because transforming a bank or credit union, because it's so regulated, it's not a trivial exercise, but those are the areas of opportunity that have to be paid attention to over time.
Michael Moeser (04:02):
It feels like AI is everywhere. Everywhere you turn around, Adobe's added AI to summarize your PDF, it just feels like every time I turn around, there's something that's being fueled by AI. Now, we've heard a lot of the stories early on about some of the failures or the misinformation, hallucinations, if you will, that have been caused by AI. And I think as I started the podcast, I mentioned there's a degree of distrust surrounding generative AI, 80%. How would you address banker's concerns about AI fueled misinformation and potentially biased decision making? Because that's a real concern?
Nima Ghamsari (04:38):
It's a real concern, and I would actually start, I would say the mistrust is probably reasonable given that one, I think people slap the letters AI on just about anything these days. And so you will hear about, oh, we have this new AI solution. Then when you dig underneath, I was talking to a customer of ours in Michigan, and he was telling me he signed up for this AI pilot, and I said, which company is it? And he said, this company X, Y, Z company. And I went and I looked it up and I was like, are you sure this is AI? It's not just some sort of basic automation and some technology, but powered by humans. And he said, well, they told me it's AI and he's not a technologist. Of course he doesn't know, but as a technologist, I look at that and I'm like, I don't think that's AI.
(05:24):
And then sure enough, a couple of weeks later, he called me, he's like, yeah, it turns out that they just slapped AI on the name of their product, and now suddenly they're pitching it as AI. And so I think for that, part of the reason for the mistrust is that where everything's AI, and so you don't know what's AI anymore. And then the other part is what you said, there are limitations to technology today where there are things that AI gets wrong. There's this thing I keep seeing on Twitter where people ask AI, which number's bigger, 9.9 or 9.11, and AI reasons. And it says, well, 0.1, one 11 is bigger than nine and therefore 9.11 is bigger than 9.9. It's like basic things that humans would obviously know that 9.1 is smaller than 9.9, but AI can't do those kinds of things. And so while there are limitations to AI, the big thing I would say is the limitations to AI are rapidly declining. And so AI today is probably about a hundred times better than it was two years ago, and it's about a hundred times cheaper than it was two years ago despite that. And so it means that that's, the AI has gotten 10,000 times better in two years. Think about that. What kinds of things get 10,000 times better in two years?
Michael Moeser (06:40):
Very little. And I would ask a question that related to the distrust, mistrust, if you will, when you're meeting with regulators, they're definitely the cut and dried type of people. And so if I stay one more minute on this, sort of the challenges I want to ask you about, what do you see as the near term hurdles that banks will face as they begin to deploy ai, specifically the regulatory hurdles? I mean, you're 9.1, 9.9 versus 11, et cetera. What are some of those hurdles you think that banks will have to worry about, especially regulators as an example?
Nima Ghamsari (07:21):
Well, I think when we're figuring out where to deploy AI, I think there are areas to deploy it where there's still high ROI, but low risk, okay, what's
Michael Moeser (07:31):
An example?
Nima Ghamsari (07:32):
So a good example of something that's potentially high ROI, but high risk is something where you're recommending products to consumers. That's a negative area. That's high risk because if AI is not there yet and it makes a mistake and it recommends a wrong product to a consumer and it hurts them financially, sure, that's not good.
(07:49):
However, one way you could completely mitigate that risk as an example is instead of putting that AI recommendation in front of the consumer directly, put it in front of your loan officer, put it in front of your banker, and have them understand it and look at it and look at the reasoning and say, oh, yeah, that does make sense. You just save me a bunch of work where instead of going through 25 different screens to find out if I can save that person money, the AI told me I can save them this much money. Let me just double check the work. That seems like a good recommendation. Let me just triple check that work before I tell it to the consumer. And so that's one example where I think high ROI low risk for the consumer and for the bank and credit union. Another example would be anytime you're doing heavily operational things, when somebody's staring at PDFs or pages in front of them all day every day, and they're checking if this name on this birth certificate matches this name on this application, okay, does this birthday match this birthday? Not only is that a really tedious and expensive action, but those people make mistakes and they probably make more mistakes than AI will make. And it's not a job that people are super thrilled to have.
Michael Moeser (08:57):
Nobody gets excited about that.
Nima Ghamsari (08:58):
Nobody is excited about that job. Another area where you as a bank or credit union, you're spending a lot of money doing that. There's not great non-AI solutions to those problems, especially ones that are real time. And so that's another area where there's high ROI and low risk because you're already making mistakes there. And in fact, I was talking to one customer about the product recommendation example, and he was actually worried about the opposite, where he's like, if I don't give my bankers and my loan officers more tools that help them understand how they can make the consumer's financial life better, they're going to make mistakes without that, he's like, we probably make a lot of mistakes today without those tools because not every banker or loan officer knows every product that's available and understands every detailed nuance of the consumer financial,
Michael Moeser (09:40):
Or they may be preferring to recommend something that they know very well.
Nima Ghamsari (09:45):
They might be very familiar with veterans products, for example. And so yeah, they might prefer a specific product area they're super comfortable with, but it might not be the best product for the consumer. So he was thinking, I'm actually probably making a lot of mistakes today, so I don't want to hold my team back by not giving them more tools that they can do a better job with. So I would say focus on the high ROI low risk areas to start. I do think that there are real challenges that AI, do think that those will go away over time, and as people get more comfortable with it, that's when we can expand into other areas that make a lot more sense to automate once we're really competent in the AI itself. So
Michael Moeser (10:21):
Let me ask you, I mean you started to talk about as examples immediate opportunities. So that's really the next question I was going to ask you is what can people take advantage of flipping that script and maybe your comparison of the driver's licenses to the application, the human in the loop in terms of recommendation? What are some other immediate opportunities that banks could think about? What could I do today to potentially test or pilot?
Nima Ghamsari (10:48):
Well, I actually want to talk about that comparison of a document to some data in a little more detail because even a year ago,
(10:58):
AI was not capable of in real time, a consumer is going through a flow. They need to provide their birth certificate, they need to provide a bank statement. The AI wasn't capable a year ago of processing that document in real time, comparing it to the data of the application and saying, does this thing match? AI wasn't really capable of doing that a year ago. Now it is possible. And so that means that your consumer can go through the flow. When you asked 'em for that birth certificate, it can extract the birthday, extract the name. You can ask the AI in English, Hey, does this person's name match this other name over here and does the birthday match? And if it doesn't match, you can in real time have the consumer correct it, explain why it's different. Maybe they had a name change, they can explain why it's different. Those are things that happen all day in banking. They happen every single day. They're happening in hundreds of other different examples where maybe the income doesn't match or maybe there's some reason that something's not perfectly right with an application for a new product, for example. And I can do that.
Michael Moeser (11:58):
It sounds, it addresses the customer experience because if it's a normal flow, maybe I got married, got divorced, changed my name, et cetera, now you're asking me in real time versus the application gets rejected, I've got to come back,
Nima Ghamsari (12:10):
Wait five days, you get pended. These are real things that happen all the time. And that's I think the beauty of it. I think you nailed it right on the head, which is AI allows you to do those things in real time, which means it really solves the consumer experience, the customer experience, whether they're walking into a branch, that branch person can deal with that in real time. They get flagged in real time, or they're doing it online, they get their application approved, or if it's rightfully denied, rightfully denied in real time.
Michael Moeser (12:38):
Sounds like it benefits. I see a benefit to the consumer experience or the customer experience. And then also from the employee experience and the bank in general to you're removing that risk. Asking someone your driver's license doesn't match your application why and feel like that you're solving that problem in real time, addressing the concerns the bank eliminates that risk or even expense if it's got to be kicked out to an exception team to deal with that matter. Sounds like there's multiple benefits there.
Nima Ghamsari (13:09):
Oh yeah. There are whole exception teams built around every one of these things because they happen all the time and because you don't want to lose the customer over that exception. And so instead of outright rejecting them, they send it to an exception team. But what if you could approve those customers in real time and actually lower the risk because you're not having a human sit and stare side by side and look at these things. And again, that's the type of thing that people don't understand. There's mistakes that happen with that today because humans get weary, get tired, they can't look at the same screen 500 times a day and make perfect decisions all day long.
Michael Moeser (13:43):
Now, I want to ask you to look into your crystal ball, and you clearly did this as a poker player. I don't know whether you were counting cards or sort of getting a feeling of how the game was being played, and you were sort of thinking maybe the big bets haven't come in or maybe you've seen a good run. But as you look into your crystal ball and you think about future AI opportunities, and you talked how it's changed from one year to today and then the potential down the road, where do you see the banking industry in five years and what will be common and what's still going to be left on the fruit tree yet to be picked?
Nima Ghamsari (14:21):
I think the thing that I'm most excited about with AI is this concept, what I call autopilot. Okay, I steal all my good ideas just for what it's worth, Tesla has this thing called autopilot, which the cars drive themselves. But wouldn't it be nice if as a consumer, when rates came down your loan automatically, once it was down enough that you could benefit, it would automatically refinance for you and get you the savings you wanted in the right product?
Michael Moeser (14:44):
That would be great.
Nima Ghamsari (14:45):
You'd save a bunch of time, you'd save a bunch of money, the bank would save a bunch of operational work, everyone would benefit from these kinds of things. Or if your money could be better allocated from a savings account or a checking account to a savings account or even to a cd, it would automatically do it for you and maybe ask you a couple of questions along the way. So one of the things that I think will happen as AI gets more powerful and as the regulation and the industry starts to understand what are good uses and what are not good uses is this concept of, I hate to call it autopilot, where the consumer will be prompted in real time when something can benefit them financially. Hey, just so you know Mr. Consumer, you should look through all these details, but this could save you $58 a month if you do this one thing. And then it'll start to optimize every single consumer's life in real time. And I'm not talking about just the problem with banking is not that it's not good for consumers, people, I think banking gets a bad reputation.
Michael Moeser (15:51):
Well, it feels very transactional. It feels
Nima Ghamsari (15:52):
Very transactional. I think part of the problem is that to get that level of financial guidance of how to optimize every dollar you put in each different thing and refinance this loan, do this thing, you need to have a lot of humans involved in that. And that's why only high net worth people get that today. Wouldn't it be great if the masses, every single person, the unbanked, the underbanked, now suddenly we're getting their finances optimized for them to their benefit. Every bank or credit union I talked to, they want do that. But you can't do that without technology. You can't really do that without AI.
Michael Moeser (16:28):
I mean, I clearly see with a wealthy person having a family office, they can pay for the extras, people analyzing the portfolio every decision. But sounds like what you're saying is now with generative AI coming up with the content developing the solutions, the bank is going to be looking more towards an individual's financial wellbeing,
Nima Ghamsari (16:52):
Hyper-personalized financial wellbeing.
Michael Moeser (16:54):
Okay,
Nima Ghamsari (16:54):
So how you hold your money might be different than how I hold my money. I may have a lot of debt that I have to refinance into a personal loan because that might be the best thing for me to save money on my monthly payments so I can get out of debt. You might be more responsible and have a home that you got at six point a half percent and now mortgage rates are five point a half percent, and I just reached out to you and says, Hey,
Michael Moeser (17:17):
I like how I'm the responsible one.
Nima Ghamsari (17:19):
You're the responsible one.
Michael Moeser (17:19):
Not toward with all the debt.
Nima Ghamsari (17:21):
I promise you that's probably also true in real life between the two of us. But I think those are the kinds of things where it'll be hyper, hyper-personalized.
Michael Moeser (17:29):
That sounds like it can create an opportunity for a competitor. I mean, if you're a large bank that isn't really employing generative AI to hyper-personalized solutions for consumers, and you're sort of sitting back and thinking about, I'm going to stratify my customers, assign personal bankers to the guy or gal that's got a million dollars and forget everybody else, let them deal with the transactional, could create an opportunity for a challenger bank or a smaller bank that really wants to look after the customer. It sounds like you could level the playing field.
Nima Ghamsari (18:05):
Well, and I think that the existing banks, credit unions, many of which are our customers, they have the advantage in the sense that they have the customers already. Their customers already get their checking and savings account already get their credit card from them, already get their mortgage from them and their car loan. And so as long as you're helping them advance their finances, whenever that next thing they want to do in life is there,
Michael Moeser (18:31):
They'll stick with you.
Nima Ghamsari (18:31):
They'll continue to work with you. And so I think it both creates an opportunity for the incumbents to take advantage of the lead that they do today have, and then some subset of those won't take advantage of it. I don't think every bank is going to adopt AI, at least at the scale that I think it should. And so the ones that don't, I think will be whether by their peers or by challengers, will probably will struggle. And so it should be interesting to see how the next 10 years play out.
Michael Moeser (19:00):
Well let's g}o back to the title of the story here, AI in Banking Science Project or Long-Term Growth Strategy. I love that idea because being a former banker, I think we always test piloted, test piloted till the cows came home, the sun went down, whatever analogy we want to use, and that's the general nature of bankers having that conservative approach with new technologies. So how do we look at that in terms of getting, people started with thinking about AI adoption and speaking specifically generative AI that creates content, creates these strategies versus predictive AI that we've had for a while. So how do you talk to Mr. Or Ms. Banker and say, okay, you're thinking about AI, you're listening to this podcast. How do you get started thinking about AI adoption and doing something?
Nima Ghamsari (19:55):
I think you nailed it when we were talking about some of the different exception cues and the work that's done that isn't necessary to do by humans anymore. There are opportunities at every bank, probably within every part of every bank or credit union. There's opportunities everywhere. And so I think the best way to get started is pick something small. This is by general way of doing things. Sure, pick something small but something small that you want to put into production and fix and something that you think can be solved by generative AI. Make sure it can be solved by generative ai. Run some test cases yourself if you want. Generative AI tools are all available.
Michael Moeser (20:28):
Something finite with a clear ROI.
Nima Ghamsari (20:32):
Yeah, I'll give you an example. We're doing with a National Financial Institution. It's around, they're using us for onboarding new customers and new members. And one of the things that they have to do for some subset of cases, they do have to collect some documentation and they have exception queue set up for those documentations. It's not exactly the use case they gave earlier, but it's similar concept. And they said, well, what if we just used AI for that one thing when whenever we required document, we're going to ask the AI the question, is this the right thing? Does this match what the person told us? And if it does match according to the ai, we're going to take it out of that queue and we're going to want to approve them. And if it doesn't match, we're going to still send it to that queue.
(21:13):
And so it's a really small use case. It's not going to make the bank 20 million more dollars or 30 million more dollars. But then you start to unlock that. And then the first thing they told me as they were working on that use case was, we do this in 500 places in this institution, right? Whether it it's a mortgage or a home equity loan, or maybe it's an address change. There's so many places where you have to collect utility bill or a bank statement or income. We do this all over the place. So if we can solve this here, I mean, think of the opportunity for us as
Michael Moeser (21:43):
Repeatable and scalable.
Nima Ghamsari (21:44):
Yes, it's repeatable and scalable, but you just have to start really small, prove it, because once you have proof that something works and is better than how you do it today, the momentum in an organization really matters very if you're a financial institution. Once you have proof and you have that momentum, it's easy to carry that tailwind to take it to everywhere else that it might be useful.
Michael Moeser (22:06):
Now, the initial opportunities we've talked about and the long-term opportunities you've mentioned as well, I feel like that there's, if I could use an analogy of mobile phones, the old mobile phones, you basically made phone calls and whenever you texted, you'd have to press multiple buttons just to get a single character compared to today's iPhone 15, soon to be 16, you've got apps that can do lots of stuff, track your health, et cetera. It feels like there's a gap between there. And so thinking about how do we bridge the long-term vision of what's possible with AI tomorrow versus what you can do today? How would you think about that?
Nima Ghamsari (22:48):
Well, I think one of the beautiful things about AI is, like I said, it's growing so fast and that allows us to almost see where things are going to go even before they get there. And so as I think about the perfect long-term view of the world, I don't think we have to know. I actually think it's just about embracing it and taking it one day at a time. It's not going to be a huge step function evolution like it was with the phones. It's probably going to be a snowball that rolls down the hill and gets bigger and bigger and bigger and eventually consumes a lot of the stuff that we do today that we don't think we should be doing, whether it's costly or turning the consumer or whatever it is. And so I think it's actually kind of nice. We get to live in the unknown a little bit in a positive way and know that it's just going to get better and not know exactly how yet. And that's pretty exciting as a technologist.
Michael Moeser (23:45):
Well, I want to use your snowball analogy. I'm sort of imagining myself in the snowball going down the hill. And to avoid that, I want to think about is there an expiration date in terms of the short-term window on AI? If you're a banker, you're thinking about using this matching technology. If I'm not doing that, is there a concern that if other people are adopting that, am I putting myself at a competitive disadvantage by not really investigating?
Nima Ghamsari (24:17):
Well, I'd say educating myself.
Nima Ghamsari (24:19):
I would say if these things create real ROI, in the short term tactical windows, if somebody is using AI to drive a more automated process like the one I mentioned earlier, and therefore their process is half the price of your process so they can onboard a new customer for $50 and it takes you a hundred dollars, I'm just making these numbers up. That extra $50 that you're spending, they can spend that towards marketing or they can spend that towards another business line to grow. And so I actually like to think about it very tactically, which is the ROI that you get is money that you can spend on growing your business that otherwise you won't be able to or investing more in AI. Going back to the snowball analogy, to then try to take that ROI and say, actually, we should probably invest even more because in two years we want to have this deployed everywhere as an example so that we have this huge, even bigger benefit. And so while I don't think you're losing out on a lot for these very tactical use cases, you're losing out on something and by not starting to think about how you can start this snowball for your organization,
Nima Ghamsari(25:22):
I do think in the long run it could hurt. Now, I could be wrong about this AI, somebody could release some AI platform that every bank, every credit union could adopt by clicking a switch because it's so smart and it didn't matter what you adopted before, this would solve all your problems. I don't think that's going to happen. I think it's going to be this more gradual snowball effect.
Michael Moeser (25:43):
I'd like to ask you about pitfalls. Now. From what I've been hearing so far, you're saying take tangible tactical steps, start thinking about the ROI, not solving for world peace, but actually trying to take one step at a time, very tactical focused efforts, and is the pitfall people trying to boil the ocean? I guess maybe what are the big pitfalls that people should avoid as they think about exploring generative ai?
Nima Ghamsari (26:15):
There's probably two that I see the most, and the first one is one thing you just mentioned, which is people try to think too big. I actually always want people to think big. I want to think big. I want to be thinking about how do I make autopilot? That's tough, but don't start with autopilot. Think about eventually we want to get to autopilot. Let's start with the first step in that direction. The very first step in that direction, the very first step for Tesla in that direction was putting cameras on every part of their car.
Nima Ghamsari (26:44):
And they said, yeah, in five years we're going to have autopilot. This was 15 years ago. They still don't fully have it yet, but the point is they started with cameras in the car, then they started collecting all the data from those cameras and using those sensors. Then they started using that data to aggregate more and more understanding how people drive. This was a series of incremental decisions that led them to a point where they can use generative AI to drive cars. Got it. And so even if you're dreaming big and thinking big, start small, as crazy as that sounds, you got to start small. And so I do see people boiling the ocean a lot, and I think that's a mistake. The other thing that I think is a mistake is because there is this worry about hallucination and because there's this risk of misinformation, a lot of people use that as an excuse not to do anything because they are so worried about what can go wrong. And what I would suggest instead maybe to do as an institution is do it in a very contained, highly overseen way. Because if it turns out the error rate was 20% when you were doing it with humans and the error rate's now 2%, yes, things can go wrong, but it's still 10 times better than the old way. And so it's actually interesting because when similar, going back to the Tesla analogy, every time there's an accident on autopilot,
Michael Moeser (28:03):
It gets in the news
Nima Ghamsari (28:03):
Gets in the news, and because people want to find a reason that, oh, see, we told you this thing wasn't going to work, but
Michael Moeser (28:08):
Think about the other thousand accidents that were avoided.
Nima Ghamsari (28:10):
Exactly. And actually there's, there's data on this and Waymo, which are actually self-driving cars in the San Francisco Bay area, they have one accident for every hundred, but
Michael Moeser (28:21):
That one per mile on TikTok and YouTube.
Nima Ghamsari (28:22):
But that one accident's on TikTok and in the San Francisco Chronicle and all these newspapers. So the people are worried about what can go wrong, but it turns out that these things can be a lot better. And I would say if you're going to do something and you're worried something can go wrong, that's okay. Just pay really close attention to it, staff it extra for the first few months to make sure that those
Michael Moeser (28:46):
Loans are, I like you're human in the loop example, giving the recommendations.
Nima Ghamsari (28:49):
Give the human in the loop, keep the human in the loop for now and figure out if it's going wrong. Because again, if the error rate was 20% before, now it's 2%. You've actually solved a huge problem for your consumer and for your organization.
Michael Moeser (29:01):
Nima, as we're wrapping up here, I'd like to have you look into your crystal ball again and reflect on what you're seeing in the market, the possibilities, and maybe provide some closing thoughts to our listeners.
Nima Ghamsari (29:15):
Yeah, I think the biggest thing I would say to the listeners is the thing to, I'll go back to something I said in the very beginning, which is in the last two years, AI has probably gotten a thousand times, 10,000 times better than it was two years ago. And it's hard to even comprehend that kind of exponential growth. The mind can't really comprehend it. And so it's something that please don't ignore it if you're going to ignore it, at least, if you're going to say no to it, at least say no to it very definitively with a full understanding of what you're saying no to, because it's something that can really be amazingly powerful in five years and can help, going back to the unbanked and underbanked thing can probably help millions or tens of millions of consumers across the country have better financial lives. And that's part of our core mission. That's part of Blends core mission. That's part of all their core mission. And so if we can find a way for it to do that, I think this will have been viewed as a huge success.
Michael Moeser (30:16):
Super. Well, that's all the time we have today, Nima. I want to thank you for providing your insights, your ideas, sharing your experiences on generative AI and the impacts on the industry. So really appreciate your time. Thank you.
Nima Ghamsari (30:29):
Thank you. Thanks for having me.
Michael Moeser (30:31):
Thanks for listening folks. And for more information about Blend, please visit blend.com. Have a great day. Thank you.
