S2 - Episode 3: Artificial Intelligence Unpacked with David Berglund, Global Head of Artificial Intelligence, FIS Global and Patricia Whitley, graduate student at Kennesaw State

David Berglund, Global Head of Artificial Intelligence, FIS Global joins graduate student Patricia Whitley of Kennesaw State University to discuss artificial intelligence, the ethics of AI, and current plus future opportunities in financial services.

Transcript

Copy Transcript Copied

Speaker 1 0:06

Welcome to the Georgia FinTech Academy podcast. The Georgia Vintech Academy is a collaboration between Georgia's FinTech industry and the university system of Georgia. This talent development initiative addresses a massive demand for FinTech professionals and give learners the specialized education experiences needed to enter the FinTech sector.

Speaker 2 0:26

This is Tommy Marshall executive director of the Georgia FinTech Academy. Welcome to season two episode three of the Georgia FinTech Academy podcast. Today I have David Berglund global head of artificial intelligence from FIS and Patricia Whitley . Uh, who's a in our graduate program related to informatics as well as a teacher and in Georgia or a professor in Georgia. Um, welcome to you both. It's great to have you.

Speaker 3 0:57

Thanks Tommy. Excited to be here. Thank you, Tommy

Speaker 2 1:01

David. Uh , let me turn it to you first for introductions. Um, love to hear about your career journey , um, your engagement with FinTech with artificial intelligence. Um, tell us about that.

Speaker 3 1:16

Absolutely. So first thank you for having me excited, just to get a chance to talk to you both about FinTech and what's going on with AI. Uh , when I look back my background, I actually grew up and I'll go, I'll go way back. We'll start way back. I grew up in , in Minnesota and I'm reminded of that today. There's a polar vortex going on and it's a chilly 30 some degrees here, even in Florida, but I grew up in Minnesota and early on, I knew always that I want to do something that was creative, that had to do with technology. And I was always fascinated by people. So it had to be something in that space. Um, I actually ended up getting an Apple two E from my uncle when I was maybe like five or six and early on that still. I remember my fascination with technology and trying to code basic programs. And I think that also kind of set me on this interesting course for applying tech later on, that happened a while later. Um, I ended up going to school in Chicago at Loyola university focused on psychology again, that fascination of people, but that was the point where I started actually getting more into advanced research, really looking at psychometrics and testing and statistics. So doing a ton of work with data. And I would say that was the point in my career. Then it was just an academic one where I realized not only can you do really interesting things with data, but the problem is often in the data and it was just countless hours of ingesting data, normalizing cleaning that I spent. Again, you could talk AI data is key there, go down the line a little bit later. Um , I ended up kind of following that interest that I had early on with healthcare ended up going into a behavioral science again, doing a fair amount of work in terms of research there. Right . I was planning to pursue a PhD in clinical psychology, so it was doing some research at the university of Minnesota. Um, very interesting, fascinating work. Um, but on the other side, I realized that the clinical work just wasn't interesting to me, I'm fascinated by people, but that just, it wasn't quite what I expected it to be. So , uh , I was able to kind of morph my career luckily into being able to focus more on technology and transformation within the hospital system. And that then kind of paved the way to get into a at Optum, which is part of United health group focused on , uh, I was a business analyst. So again, very focused on data and analytics for the company. And there was many just amazing steps. If you ever looked at my resume and say, well, how does this make any sense if you look at all these dots, but it was one good step that led to another where I was able to kind of join in and eventually build this innovation group within United health group. It was really a round table and folks who knew that they wanted to do more, more creative, more impact, and we would brainstorm new ideas, actually pitch them to a venture group within the company. And it had just all sorts of fascinating opportunities to kind of pitch new transformative ideas that led into me getting an MBA at university of St. Thomas , uh , focused there on strategy and my background. Again, I didn't have enough experience at the time with business. I knew that , um, but I was able to fulfill that through that education and that along with some other opportunities, open doors as I'll call them, allowed me to then become one of the first entrepreneurs in residence at United health group, where we were able to build essentially startups with internal venture capital , uh, for the company , uh, go down a little bit later down the line. And , um, not only could I expand kind of, kind of leading a number of the different entrepreneurs , uh, but eventually we kind of morphed that slightly into the emerging technology space. So I was the owner of AI within that group for advanced technologies. And obviously that then became what allowed me to kind of transition to where I've been in that space for a while now in the world of AI. And at the time the focus was really , uh , natural language processing. There was some modeling that we were doing in terms of kind of different predictive models, but a big clear focus for us was the problem of the efficiency. And we'll talk about this later, but just trying to help people who have a question about their benefits and asking questions, how can we make that easier for our employees, for our clients make that easier overall, go down the line a little bit and , uh , another open door open and I got a chance to join another local company. It's Minnesota, a us bank, which is tremendous , uh , uh , job that I had there. So I was the head of artificial intelligence there working within the innovation team. And funny enough, I had actually been doing consulting , uh , for , uh , a number of firms and ended up beating Dominick Ventura ,

Speaker 2 5:58

Dominic. I work on Dominique's team. Exactly. Yeah. I mean , everyone knows . Um , I used to interact with that team. I did a lot of consulting work for Martin Jordan falls , welfare organization. And, and then , um, there was a couple of different threads that we had that were intersecting with , uh , Dominique's team , um, which was , uh , their grad air . Just what a great team really . I have a huge amount of respect for them.

Speaker 3 6:26

Absolutely. So that was just a wonderful opportunity to connect to the dots. Just almost seamlessly from the work we were doing with startups and innovation and technology, it was the perfect fit. So truly just had a wonderful time. There did all sorts of really exciting things with Dominic and us bank. Um, and it was kind of around that time as I was again, moving from healthcare to financial services, big jump, obviously. Um, but also got a chance to , um, start presenting. Actually I was an adjunct professor, still am at Wharton where I focused on risk , uh, within , um , financial services and really a focus on AI there. So couple, you know, a while later I eventually got the great opportunity at another open door and here I am that FIS and I've been here for about a year and a half.

Speaker 2 7:14

Awesome, great. Thanks for all that background. And it's nice to, it's nice . It's cool to hear the career that our shared connection with dominate ventures group . That's neat. Um, Patricia, tell us, tell us about you.

Speaker 4 7:30

No . Okay. Well, it's interesting, Dave and I have similar , um, connection points in our backgrounds because I haven't heard care background . Um, and then I , um, I will just start mine after, when I went into graduate school for my first time. And that was at the university of Georgia. I have a son there now, so go dogs. Um, and I got a degree in organizational psychology and communication , but from there I started working for a company called digital equipment corporation. I don't know if anybody is still around, but recognize DEC . Yes, yes. And so I think the important point there that I would make and I , you can hear it in Dave's story is that I didn't have a technology background, but I went into a technology company , uh, ended up being a , um, manager of some of the best technologists I have ever worked with, stayed there for a decade and then joined various , um, as technologies changes changed, I left , uh , there and then ended up going with a client server consulting companies, et cetera, et cetera, ended up working for a venture capital company called technology ventures. And , Oh , you know, do you know technology ventures ,

Speaker 2 8:47

Uh , it's the same, we're talking about , uh, Sean Banks and that crowd TTV or a different one. No, this

Speaker 4 8:56

Was , uh , this was a different one. And Joe McCall, does that ring a bell too ? Joe McCall was the guy who was the principal there, but nonetheless experience was great. And , um, so I worked there for a couple of years and then , uh, joined another company guys might know about is the accounting firm that I was with, which was grant Thornton . I went to grant Thorton was their business development manager , um, and then joined medical device company. So in the medical device company, I was in marketing. So I think the thing that is really important is that you can, if you can learn and have a lot of interests, it really can just move into different spots and positions within companies. Does that make sense different? You don't have to actually get a dad as long as you ,

Speaker 3 9:48

And I think that's how the best things happen . Both companies do it or get organically. They realize where the puck is going, how they need to transform and individuals do the same thing. At least the ones who , who spot the opportunity and are often willing to take the risk or at least learn. Um , I think that's a great path.

Speaker 4 10:06

I think the point that you just said is the most important, and that is if you can just learn and if you're ethical, you have the motivation. So I got into where I currently am because the medical device company that we, or , uh, that I was with was working with biophotonics and we were working with , um, tumors and pre if we could come up with a device that would indicate whether or not the tumor bed had been cleared of , um, cancer cells during its first incision. So you didn't have to go back if they found tumors during pathology. Um, so when they ran out of funding, I joined , uh, Gwinnett tech to teach there , uh , and started to company or started working in a program that combined my healthcare , my science , uh , teaching, and then my technology background. So that was a joint venture from Gwinnett tech and Georgia tech. And it was a healthcare technology certification. So I came into Kennesaw because Kennesaw had a masters of healthcare management and informatics. So I was getting a degree still am. And this is my last semester, so that I could know more about what I was , uh , the lead instructor for now, unfortunately, that program closed at Gwinnett tech, but I've stuck with the program, the master's program. And I think I'm just a student of everything that Dave is the master of when it comes to AI. So I'm kind of very interested to hear where he's going and how, what, I'm just now learning. He's got the expertise in. So

Speaker 3 11:46

The beauty of it is it's all changing so fast. I think we'd be fools. If we didn't admit we were all students. And I think that's the fun part is, is the things we don't know that is often what's going to be the most exciting part of our future. Right. And we can barely explain some of the most complex things today. So just start to imagine what we're going to start to do in a few years. And it's because I think it's that curiosity like you, what I heard from your career was this curiosity and willingness to take a step to do something different. And the same thing is with technology and , and Kevin Kelly, he's the author of wired or executive editor of wired, and we're working some really great books. He talks about this idea of the inevitability of technology. And I feel like in some ways that goes down to what you were just saying, Patricia, which is the technology is almost unfolding before us. We just happened to be the ones who get the chance to work with it and apply it to, to business.

Speaker 4 12:45

Hmm. Yeah, that's great. That's great. I think a lot of people have said to me and about mine and their own careers that sometimes like just having faith walk into the abyss or fall into it, and that you're going to be a , you're just going to fall on a ledge . You'll work with it and then move on. So yeah, that's a great, that's a great attitude to have ,

Speaker 2 13:07

Uh, I love this topic as well. I was fortunate. Um, and my last role at Accenture , um, to be part of a , an artificial intelligence , um, initiative that we were working on Excenture wide and I was representing like the financial services industry track. And , um, it was , uh , you know, in Microsoft and Greg always kind of gravitated towards the emerging technology , um, opportunities. And I've always found and loved the fact that financial services is a frequently a leading industry and embracing , um , emerging technology. Um, and what I began to discover in that, in the role I was in at Accenture was , um, just how broad of a topic artificial intelligence is. Um, really, as we were looking at it, there, we were in compensating, everything from different aspects of automation all through what you'd mentioned, natural language processing earlier through that whole zone, all and done on into machine learning. Um, and these more kind of predictive aspects of the activity. But , but , um, David, can you kind of paint that picture for us of like how you view , um, artificial intelligence today? What, what, what is that as we define it , uh, today?

Speaker 3 14:38

Absolutely. And the beautiful and challenging thing about AI is we can all be talking about it and I'll be practically talking about something different and there is no, I don't think there's a singular term and I don't. And I think that that's okay. Um, when we talk about AI and I guess when I talk about it, I think of it as this kind of set of definitions, ultimately it's the ability to, to predict or, or create an action , um , using software technology that's based on data, not predefined rules like we would have done in the past. And the abstract view is it's basically being able to do the things that we only used to think a human could do with technology. I think that's a fair way of putting it and we happen to use a few different mechanisms to get there. Right? There's I would say three general technologies within AI that you would look at. One is the NLP natural language processing space that has to do with text. One has to do with computer vision, which is clearly blowing up these days. Um, both in terms of, you know, facial recognition and things of that store, as well as you look at what's happening with self-driving cars like Tesla. And so on, right. Third category is I just think generally predictive models, right? They can be identifying anomalies, it could be forecasting something that's going to happen in the future, but those are the three buckets that I think about.

Speaker 2 16:00

Yeah. Um, that, that makes sense to me. Um, and I like, I , and I , I like to think about AI in that way too. I guess what, what what's left out of that definition, I guess just in the experience I've had, maybe with us for five years is this area of robotics process automation, which I know when I first began to engage with kind of the umbrella artificial intelligence topic, RPA was included. And as I began understanding RPA more, I was like, why is this included talk about artificial intelligence? Because I was fighting with robotic process automation. It was very much okay, we've got a , uh , an application or that can just mimic what humans do on like basically keystrokes or clicking on screens. Um, and so I guess since I , so I guess I can understand, okay.

Speaker 3 16:57

I think that it's, it's a great point. And honestly, this is something that I've, I've looked at in various ways throughout my career, looking at kind of AI in that space. So right now the intelligent automation team is on my team too , as well as we've got a corporate data team. And we've got another kind of it automation ops team, and you could easily make a case, well, Oh, those are different categories. But the fact is strategically and foundationally, the goal is to drive business value using this suite of technologies. Right. And along the way, as we apply RPA, whether or not it's UI automation or we're hooking into API APIs, the goal should be to enable us to do more advanced work, to , to unlock the machine learning through improving the data connectivity of systems. Yeah ,

Speaker 2 17:45

Yeah, yeah. That makes total sense to me. Um, and that was how, I guess I was starting to think about it too. I was like, okay. So it kind of like the RPA ended up sort of being the training wheels place to get started. And, but then I could start to see how it was, is vital, particularly at accessing data, drawing data out of some really difficult places where it can be very difficult to get to in a, in a very efficient way to then be put somewhere so that the more addictive , uh, elements can be

Speaker 3 18:20

It's the action piece too, right? When people think RPA it's all action oriented, you're trying to do something. And I think that's often the missing piece for me with AI, as many people just think of it as a model. Well, the model is great, but only until you apply it to a user experience or a business process, can you make something change? It's that action that creates value and improves an experience?

Speaker 4 18:47

Sorry, can I interject something real quickly? Well , you mentioned that I taught at another university. I just wanted to say that one of the papers my students write for me is will robots take my job and the , uh , I I'd love to hear , um , just , uh , Dave's comments, maybe yours too quickly on , um , what my students are thinking through. I'm asking them to think through, and that is, will robots take my job? That's the title, but how will their future careers interact with AI and robotics? And I often show them the Atlas videos from , um, Boston dynamics, et cetera. So any comments there about how do you think that's all going to work? You Tommy brought up robotics. So it took me right to my student assignment.

Speaker 3 19:33

Oh, it's a great question. And what a wonderful question to ask students. And I would say there's probably a few different answers to it. The shortest answer, the shortest paper you're going to write is yes. Right . That if that's all you got back, you could probably give them

Speaker 4 19:47

That's an a, okay. That's a difference . I think it depends. Yeah .

Speaker 3 19:51

The timeline, right? So some jobs are inherently more challenging to replace and many, we don't want to replace them . It is the human connection that I think we need that we're not going to be able to replace anytime soon with AI and think of that as many jobs in the service industry where people appreciate that connection with people, or that require creativity, judgment. You know, some of the things that people would say are the best of humanity. If you also look at it in the near term, there's a lot of jobs that are dull, dangerous, dreary that you could replace with AI or automation. But that doesn't mean that that is the pure intent. I think we should then flip it and say, how can we help enable those people to do find the next job where they can add value? Because surely there are those opportunities. And this is something that I would say many companies struggle with is the change management to people. How do you get them on board to help them realize it's not to be afraid. You don't have to be concerned, but rather let's help them get enabled to make the change and figure out what they can do next with more impact.

Speaker 2 20:58

Yeah. And this might be just a moment for us to move into this broader topic around the ethics , um, around artificial intelligence. Uh, cause I think, I guess I agree with you, Dave, there's the answer to the question is yes. In certainly in my lifetime. And so, but, and I think then it's okay if we understand that if we believe collectively that that answer is yes, then there's an ethical responsibility that we have knowing this, or believing it as individuals and as companies and as governments and as schools to start figuring out that type of trend , that those sort of trends that transition that you were mentioning where, you know, what, what does that look like? Where is this appropriate and getting that , um, starting to investigate that and coming up with solutions , um, in really every different iteration we can begin to.

Speaker 3 22:02

Absolutely. And it's a tricky one, right? I mean, ethics is something that we all can recognize is important for organizations and people to just think a lot about. And we also can likely recognize that no two groups or no two organizations may completely agree, but that's where I think, especially with some of these technologies, whether you call them emerging advanced technologies, they can and have shown that they can play such an impactful role in people's lives in all of our lives that we have to at least think about it. We have to put together some basic principles just as we would as an organization or a government about, you know , what is proper. And what's not,

Speaker 4 22:43

I was curious about just , uh , a definition of ethics , um, at this point, could one of you, or , uh , could we add that definition in their ethics ? I think sometimes it's too loosely defined. It'd be nice to have a working definition while we've talked through it. I would agree.

Speaker 3 23:01

Um , I'm definitely not , um , an ethicist myself. Um, so I probably can't give the most formal definition, but I, I think it , to me, it comes down to the morals and principles that we agree we ought to live by. Um, and , and , uh, maybe I'll keep it simple, you know, for FIS one , three, the three kind of key pillars that we look at is how can we make a positive difference for our colleagues, communities, and clients. Right. And we keep it that simple. Right. And then when we think about applying AI or any other technology, we can do it through that lens of what , who are those key stakeholders and how can we make sure we're making a positive difference for them in, in a meaningful way?

Speaker 2 23:48

Um, right. Exactly. I mean, I agree with that too. Um, and I'm, I'm neither am I an emphasis? However, I am married to an SSS and I'll get her on the phone if we could , uh , yeah . If we could get her right now, she could , uh , really give us an eloquent , uh, answer to that. Um, but I , I, I, so I think it's that , there's the, you know, how there's a responsibility to , um, us as you know, involved in the evolution of this technology to be making sure that it's serving the advancement of us as humans and not , um, and not creating , um, a, a kind of holding us back as humans and as our , in our human evolution, I guess, in a broad stance, a very conscious

Speaker 3 24:48

Sizeable view, right? That's thinking the biggest picture from a humanity humanity standpoint, which I think is the right place to start. And I'll also say that if you approach it at that level from many organizations, they'll think that you're almost imagining a far off future of science fiction and that while I think we should, we also have to translate it into what's a real impact of today. And we're seeing them right with how, whether it's social media companies or advertising or pick , you know, a large organization that is able to use data in a way that influences people, people are starting to recognize that their data is being applied often for good and often for ways people are uncomfortable with. Right. And I think it's the , the times where you don't see, you don't expect what's going to come around the corner, the unexpected ways that the technology can unfold. Um , the dangerous ways that becomes to me, something that we should really think about is leads to have a checkpoint, you know, how could this go wrong in the future?

Speaker 2 25:50

Right. Well ,

Speaker 4 25:51

Let me interject to that. I think from my standpoint, as both a student and as a professor that you have, my definition of ethics is are we appropriately training students to enter into the world of AI and to robotics back to the original point of my question . So , uh, or the paper that my students write ethics, meaning are we actually preparing people to enter into this world appropriately? And that goes back to what you were saying about when robotics, people are seeing their jobs, or maybe filling the threat, how do we teach them to either move into something where they can bring the value of social connectedness, which I think is important for human interactions. We don't want to lose that. Um, it's just , um, I think being conscious and that to me is , um , ethical about how we prepare people to work and with technology. And so that I think is something important just launching students into their careers from the beginning. Do you know what I mean?

Speaker 3 26:53

I , I completely agree. And this is where , um, there's the, you know, the Spiderman quote with great power comes great responsibility. I think that's what some of us should now recognize we have. Right. And, and you could also say just because we can do many things, it doesn't always mean we should, and then you can easily start to say, okay, well, there's a couple of factors there that we need to kind of unwrap. One is we need to be able to have some level of explainability in the things that we are creating and putting into the world. So people can ask proper questions because if you can't understand what's going in and how the model is making its decisions, you can easily start to contribute to the unfair or biased treatment of people. And obviously I would say, hopefully, obviously none of us want that. So that's just, you know, keep it simple so that someone can understand it and ask the appropriate and proper questions. Um, so we don't continue to foster biases that may have been in , in the world and in the past that we definitely want to avoid in the future.

Speaker 2 27:57

Yeah. I want to push on, that's a great point, explainability, and I want to then bring us back to financial services and specific use cases , um, that, that you think are really important in front of us here in the next , right now in the next couple years. And then to the extent you can share a bit about , um, you know, what your , what, what FIS is working on, that would be great. And so , so I'll start with the explainability point is one that kind of first I started paying attention to in the context of , um, credit underwriting in financial services. So I think many of us see a , a really incredible opportunity to expand , um, access to credit, to more and more humans, not just the wealthy , um, as a result of artificial intelligence. And so that the way that kind of generally works is if we can get more data on individuals than a credit underwriter can understand more about that individual. And then as a result of that, understanding is able to manage their risk, which they need to do appropriately if you're making loans, particularly on secured loans. Um, and so explainability of course becomes important because you need to be able, if you're an underwriter to show your board, to show your executives, to show your regulator, this is how precisely we've made a , um , decision related to lending this new person money. Um, and , um, there's still, I mean, I think there's been great progress in regards to explainability. There's still more to be done in the context of , uh, of credit underwriting, but , um , that's just a comment for me to kind of, you know, if you want to respond to that, but then talk a bit about other use cases you think are really important to be focused on here in the near term.

Speaker 3 30:09

Yeah , I would say you're spot on. And that's one of the ones that is hits people from both a what an amazing opportunity view as well as a well, but wait. Right. I think it's fair to ask both questions in there somewhere. I think most of us could agree that AI being applied and, you know, anti money laundering or fraud, like, of course let's get the bad guys, you know, find a way to get the bad actors out of the system. Many people would be comfortable with that. And even if we look at it from a kind of transactional standpoint, how do we enable a more efficient and improved experience with banking, right through chat chatbots , or virtual assistants and even personal financial management, all sorts of really interesting and positive things can be done there. But I would agree with you that that space of credit decisioning for all the right reasons, there's reasonable scrutiny, both because it's interesting because all of a sudden, now you can help the thin file. No file customers get access to credit, but often to do that, you have to think different, like the system has been built around the model that we've been using for a long time. So how can you incorporate a new type of data, a non-traditional piece of data and help someone understand how that contributes to your decision to give them credit. And, Oh, by the way, your point earlier about, you know, you've got all these groups that you need to be able to explain the credit model, too . Of course we do. And the person that we also need to be able to help inform is the person who is making the request, how do you help them understand why they were rejected ? And now you can't just say our model said, so you have to be able to find it, I think potentially ethically help them understand how you can get them back on the right track so that they can be approved in the future.

Speaker 2 31:48

I love that . And then how about , um, you know, what's, what sort of David Wright kind of generally in front of you right now are the problems that are really exciting? You , um, yeah. You know, I think about FIS global, I mean, certainly one of the largest financial technology providers in the world , um, the, you know, as a result of that tremendous opportunities, tremendous access to data, tremendous resources to bring to bear. I mean, they'll , I mean, there's just no question, like you're in a position to be a real leader in industry-wide , um, on, in the, in the engagement of this technology into financial services.

Speaker 3 32:32

Absolutely. And I would agree, and FIS has done a tremendous job for a long time, as well as some of the organizations like Worldpay. You know, that's part of FIS now doing interesting things, not only because they had to, to stay competitive, but because they saw the opportunity. And a lot of the things that my group has focused on have been in that realm of how do we help FIS run more efficiently and how do we help our clients, the , the bank and banking institutions who we work with and help them understand how to get the most value possible out of our products to then service their customers as efficiently as they can. One project that we're working on that's public now , uh , to some extent is the opportunity to work with the FDA on improving the way that they can understand risk in the banking system . And like you said, we have a whole host of data within FIS. And the question is actually kind of comes back to before, what data are we allowed to use? And we're very thoughtful about that, but how can we do it in a meaningful way for the FDA so that they can spot systemic banking, risk , faster using more advanced models and more granular data. And it's something that they've been clear, you know , about publicly that does is what they're looking for. And we're well positioned to solve it both from a data perspective and analytics, and it's not just meaningful for the, see , the thing that I'm most excited about there is how do we help avoid the next black Swan event, the next, you know , potential significant downturn from a COVID like situation or another kind of housing crisis situation by giving those insights, not just to, you know, the, you know, our , our regulators, which is the right thing to do, but all the way down to the bank so that they can spot things faster, the things that they may not have had the tools, the technologies in place to do before, and with this holistic view, we can help make sure that they can find a way to fix that trajectory as fast as possible.

Speaker 2 34:26

Yeah, no, that's awesome. Well, Patricia, any ads from you or are kind of even beyond financial services kind of use cases that you know, are really , um, meaningful ones to focus on in the near term?

Speaker 4 34:41

Well, I think the thing that I done the most research on it has been dealing with heterogeneous data. Um, I don't know what day, but to think about this because he's a practitioner again, as we talked about where I'm a student, but pulling in information or trying to get information from , um, like EKG , um, and text, and you're dealing then with just , um, what I might call Excel, like data Excel, database data, but he just got , um , the common patient ID and things of that sort. So heterogeneous data would be one place where I can see AI trying to make quick, quick use of large data sets that come from so many various sources. Um, and then I think the other thing that is concerning, because we're always worried about this and healthcare and that , uh, privacy and , um, HIPAA violations or avoiding type of thing. And how, how do we pull information across large and various patient data sets without , um, well, I guess maybe with staying inside a de-identified data, does that do that ? How does that register that ring true with anybody's U2 at all?

Speaker 3 35:54

Yeah, absolutely. And I would say I was going to touch on what you mentioned regarding the heterogeneous data. Something, when I think of there is how do you combine different data sets that people used to think they never belong together? And partly it was because we didn't have the place to put them right to the database structures. They weren't possible to put them together, but also we didn't have the systems, the compute, the power to actually find the insights, the correlations and the relationships. And if you look in there's , you can easily make many comparisons between healthcare and financial services, but on the healthcare side, how can you look at the unstructured text from a doctor's note, along with all of the biometrics, maybe within labs or the, you know, temperature, blood ox , and all those things along with an MRI. And it is across each of those completely disparate systems that you may then be able to help spot , um , potential cancer or some other looming condition that we never could connect the dots before. And I think that there's the same type of power that we can apply to loan applications, right? We've got the same similar unstructured documents and PDFs that, okay, you start extracting that with computer vision applied some interesting models, normalize the data and boom, you can help someone get on a better path to get credit and improve their finances.

Speaker 2 37:19

Yeah. Cool. I want to transition us to talking about recent news FinTech news. That's caught our attention , um, in the last week. Um, David, how about you? Anything caught your attention,

Speaker 3 37:35

The big one for me, and I've been trying to better understand it is the , the whole I'll call it the Robin hood side of things with GameStop and, and GameStop has been on a life of its own, just in terms of all of the, the press, the questions, the attention, and what's happening with wall street bets. And I just find it fascinating. It's not an, an AI play per se. I w I do think it will be interesting how different big hedge funds start to discover this a little faster, maybe with natural language processing. But what I found interesting was Robin hood in the attention there in terms of the , um, what the data that they have and that they're potentially giving to hedge funds and others for within the order flow and the questions they're being asked. I think that'll be interesting. What do you think about that?

Speaker 2 38:25

Yeah, for sure. Well, I guess I'd say this last couple of weeks with Robin hood , I think it's brought some much needed an important attention on Robin hood , which, you know, some would say has been kind of skating on thin ice for quite some time. Um, you know, they launched whatever three or four years ago, they've acquired a lot of customers may lead gen Z customers , uh, with their zero fee attention. And then the real kind of one problematic, I think, aspect of their business model is there's quite a bit of game of cation technique that they use to encourage trading. Um, and then then additional another problem that had them is what you're bringing up, which is their entire business is focused on what's called payment for order flow. And so , um, there is spread in every single trade and those individuals, any of us that create on Robin hood are two , are certainly being taken advantage of and not getting the most appropriate pricing versus those that trade on. Um, perhaps other other platforms , um, or, you know, I'm saying that as factual, but this is , um, you know, been , is often in discussion , um, in the marketplace when it comes to this payment for order flow business model circumstance. Um, so, and then certainly , um, the, just the game stop mean phenomenon, you know, largely got traded by the wall street bats crowd was largely using Robin hood , um, for their trading there's other platforms, of course, involved you trade and trial .

Speaker 3 40:07

And the power of the data, I think comes in here, both the power of the crowd to come together and kind of centralize on the, on the power and the questions that they were asking from the Reddit side. And then the power of the data that Robinhood I think was really probably in a really intelligent way, able to recognize they could create a whole new business model in some ways, using that data to unlock access for a different set of customers that maybe couldn't before with how the system works. So it's interesting, I'm glad people are at least understanding the dynamics of how it's, how it's working.

Speaker 4 40:39

Um ,

Speaker 2 40:40

Patricia, how about you, any neat news caught your eye?

Speaker 4 40:44

What about that? Uh , JP Morgan chase launching the digital consumer banking in the UK. Why did you think of that?

Speaker 2 40:50

Right? Yeah, that was , um, I thought very interesting. Um, so this is , uh , just UK fo United Kingdom focused. And , uh , you know , if you look at that UK market from a consumer vein standpoint, there's been a real , uh, for years, this isn't a recent phenomenon, but for years, there's, they've really embraced these kind of either web first or digital first plays more recently it's companies like Revolute Monzo , um , in particular that have just been, you know , adding millions of millions of customers. And so I think chase is seeing they are an opportunity. Um, some of you may remember that they had launched a consumer digital play in the United States called Finn F I N N, that they shut down or didn't shut down. They folded it completely into the U S chase digital app. Um, so they kind of had tried a similar play U S focused that wasn't materializing as they did. So they folded into their , just the normal chase mobile app. Um, so I think it'll be neat to see how this , um, kind of chase going to head to head with the revolution Monza place in the UK,

Speaker 3 42:03

You have different advantages, right? What can they bring to bear that maybe the competition couldn't, and in the same way, they're going to have different struggles. Um, this is new for them in some ways. Sure. They've got some of the things that they've done with fin , but it's a different market, different expectations, and how can they adapt in that way will be really telling and in the same way, how do they pull back the insights that they pick up through this? We'll call it a test. I'm sure it's more than that, but what they learned from applying this in the UK, there's probably a lot that they can do with similar audiences here in the U S

Speaker 2 42:35

Yeah. Um, and then the last news story, I guess I hit on is this ant financial , um, news that they're now going to be , they're essentially being forced to restructure as a holding company. Um, and that's by the Chinese government because the Chinese government really, I think, has concern . They want more, they want to have better insights, more direct access to, I think, data up . They're not saying that directly, but I think that is an intention. And , um, I think that's an important , um , development for us, I guess, as global citizens, but then certainly in the U S to continue to keep , uh, some focus on. And there was a really interesting 60 minutes , um, CBS, 60 minutes over the weekend where they were digging deep into China's kind of broad statements about, we want to be the global leaders when it comes to , um, data accumulation, consumer data cumulation, and, you know, leading with artificial intelligence. And that, that story was 60 minutes was very focused on bio bio. I don't know I'm using the right term, but it's like the biological bio-related data. Um, and how China is. There's a lot of different mechanisms they're taking to try to just obtain massive, massive, massive, massive data sets , um, related to , um, every individual on the planet. Um , hoping to then I guess, create products, data products, where they would approach any of us and be like, Hey, would you like, you know, pay for X and we can help predict better what your health outcomes are going to be in the next five years. I don't know, but it was really,

Speaker 3 44:20

That makes sense. Right. And you think about it from a national perspective for advancing AI, right ? A key has to be how do you access as much data as possible? And if that is a move with ant that allows them to get that access that puts it , you know, put someone at an advantage and then you could easily ask for any company, organization, whatever you are, is that proper? Well, the question to me is how do they use it and think of the good that they could do to use . And you talked about how they could apply with health, same thing with COVID. How could they have some of those insights with purchasing behavior banking data as a way to help reduce the spread of COVID. That could be pretty powerful. And I think some groups have proven that there are trails, breadcrumbs, whatever you want to call them that could help us figure some of those things out.

Speaker 2 45:10

Well , it also takes us back to the slippery slope of ethics too. So , um, I'll, I'll just leave with that. Yeah, no, absolutely. Well, I'm going to wrap there, David Patricia cannot. Thank you enough for being part of this conversation today. Thank you to FIS global as a investor , a founding investor in the Georgia FinTech Academy , um, really deeply valued that relationship, that partnership, and greatly appreciate it. Uh, and you're, you're both welcome back any time . And , uh , thanks for your engagement with the Georgia FinTech .

Speaker 1 45:46

The Georgia FinTech Academy podcasts are available on iTunes and Spotify to obtain additional information about the Georgia FinTech Academy. Please visit our website@georgiafintechacademy.org.