Monica Millares: [00:00:00] Hi Paolo. Welcome to the show. It is a very exciting conversation that we're about to have.

Paolo Sironi: Marika, thanks for the exciting invitation. I'm happy to share with your audience.

Monica Millares: Thank you. And I think just for everyone, Palaboy just published, what I would say it's a book. It's not a report. You can see it's generally a book about AI and financial services as such.

Here we have it. And what I really liked about it is that you can see a ton of short reports online today. But it's, sometimes it feels like it's just a lot of noise or it's just like repetitive and you feel a little bit lost. And what I liked about Paolo's report is that it is all in one page, very not in one page, all in one document, very structured.

It follows a nice sequence. It's easy to understand. It has case studies. So I [00:01:00] thought it was worth going deeper into this conversation so that we can complement the book with your explanation

Paolo Sironi: as such. Thank you. Effectively, this is actually the 2024 Global Outlook for Banking and Financial Markets is a series that is published every year from the IBM Institute for Business Value.

I am the global research leader in banking and financial markets, and I tend to discuss with my colleagues the economic models of financial institutions and reflect upon the role of technology in that transformation. So this year, that reasoning is intersecting AI and generative AI. In the background, we always have the economic model of banks, because technology makes sense when it is applied to a business function or a business need.

Otherwise, discussing as abstract will not add value to anyone in the industry.

Monica Millares: Exactly, definitely, [00:02:00] before we go in detail into the conversation, this podcast is about how we can build better fintechs, more purpose driven fintechs. So what's your opinion on how we can build more purpose driven

Paolo Sironi: fintechs?

It's a very, important question. And the reason it is, because I've been also doing a business literature in my professional life. And the latest one banks and fintech on platform economies, starts from a foundational question, the question of value. What is value? If we don't define value, we cannot digitalize value, and I feel like many have distorted or not perfect understanding of what value is for the final consumers, first of all, and then out of that, how can we position?

Technology and business opportunities, of course, to do a business, but bearing in mind that what matters is to create more value [00:03:00] for the individuals. And what I also care about is that value is not remaining at the level of the micro interaction. So FinTech and consumers, banks and clients, but can also expand to the level of the ecosystem, which can be society or the economy overall.

So I've been researching on a theory of value. My financial market transparency theory principles is meant to do that looking at the platform economy to understand how in transparent markets, technology can help create value for the individuals in a way that it basically propagates a sort of spontaneously throughout the ecosystem is so complex these days to have a perspective, which is it.

Open them, but at the same time allows to reconcile the privacy of individuals with the need for an entire ecosystem, basically, to fulfill the broader society goals and scopes. And so I think that I'm addressing this in my literature. [00:04:00] So value is a must be at the core. We can have a slightly different definitions, but starting from there, we basically understand that where we go to, and I think that this paper as well, the regenerate banking with AI has that somehow embed that in the core because it is a story about humans.

People, employee that interact on the platform economy with institutions, regulators in a way that the interaction between people and technology can basically generate that more value for everyone.

Monica Millares: Yes. And then I'll just go straight to it because exactly the keyword is value. And I love you're the very first guest that talks about value and the definition of value, because if we don't have a definition, then I cannot generate value.

Paolo Sironi: Yes. So now we may have to explain maybe in another podcast about this right there, but if you think [00:05:00] about for example, what banking can do through technology. Is that to eliminate their friction in client journeys, and there may be different friction that can be realized and different client journeys.

One of the simplest one can be, you pay one click when you're on Amazon, basically use technology to enforce payments, which are instant secure. And, added value for clients who don't have to otherwise spend time in doing that element. But there's deeper value that is part of the industry that can be addressed that also looking at technology.

And that is the, basically there's a knowledge friction. The majority of people as have a hard time in making a financial decisions, which is the reason why banking is a typically based on human relationships. And that's the reason why, even though the majority [00:06:00] of consumers also reported in this report currently like to engage on digital medium, like mobile apps, looking at their transactions, accessing their bank account.

But then in the end, we also learned that. It's one thing is executing a decision or one is making a decision. So execution, a lot of people are asking for digital, but making a decision, typically look at human relationships, which are complex, expensive, and also bankers needs to be elevated. So now if, the foundational value of the industry, Is to share knowledge among the parties how do you eliminate the knowledge friction?

When you respond to this question, you unlock the biggest value that the industry through FinTech can basically generate. And from there I start in particular in my literature to discuss the transformation of platform economies, of advisory networks, and so on and so forth. So that's where I'm basically lending to going to the core of the problem to enable people to [00:07:00] self direct themselves transparently being in control.

And, being part of a fair treatment when they deal with FinTech or financial institutions.

Monica Millares: Yes. And then I love that you talk about knowledge because now it's been a year, right? Since AI became Ooh, the buzzword, but now it's not a buzzword. It's a reality. And I think as a FinTecher, I think we as a community, some of us talk about AI, but we.

I really don't understand it and many of the banks or the fintechs, we don't have the capabilities because we're banks, right? We were not AI technology companies. We are building the capabilities. So therefore these removing the knowledge friction, I love that concept, removing knowledge friction, it becomes so relevant when it comes to AI.

Paolo Sironi: Monica there's one message inside of these global outlook [00:08:00] for banking and financial markets that matters in this conversation. I started my career as a risk manager. I was head of quantitative risk management in capital markets for 15 years. Then I built a fintech that IBM bought, so I joined IBM some years ago.

But I remember in the 1990s, where risk management started leveraging technology to create architectures for Basel 1, Basel 2, so complex calculations and aggregations of analysis and reports, that risk managers had this mantra. They said every bank employee is a risk manager because risk management is the core of a financial institution.

Now, today, I think we need to say every bank employer must be an AI risk manager is so important that we understand how AI works, the advantages, but also the risks and the complexities that because we need to leverage this technology to transform the bank. And this is not a battle human machines. I don't think I will replace.

[00:09:00] Bankers, but bankers who are capable of using AI will advance compared to those who can't. So everyone is made responsible for their own benefit to basically become an AI risk manager. Definitely.

Monica Millares: That's a good message. So now let's go full on into the conversation, but before I do just a little bit of context, basically we go back.

You just released the your latest research that it's the global now for banking and financial markets regenerate banking with AI and you surveyed 600 banking executives basically on how they are using generative AI. There's you added like some stats at the beginning of the report that says almost eight in ten institutions are implementing generative AI for at least one use case.

And basically what you just said, AI basically reflects concerns about risk and client's relationship and governance is a must have. So [00:10:00] before we go deep, deep, deep into this conversation, can you expand? What is the one takeaway that we need to take? To think about. And I think you just touched on it so we can

Paolo Sironi: go deeper.

Something very important. Of course generative AI is just one part of AI. So the conversation could be broader in this global outlook. We focus more specifically on generative AI being the latest and greatest. So when it comes to Generative AI, 86 percent of financial institutions worldwide are live or implemented to go live with at least one use case that uses Generative AI within risk and compliance, IT development, other support functions like finance, auditing, HR, as well as client engagement experiences.

But 8 percent only, I would say only of those have a systematic approach. That means they are working to implement this [00:11:00] technology across all of the business domains of a financial institution, while 76 percent of the remaining gotta be more tactical. That means that they are working on one use case here, maybe another use case there.

So there's not a single story that dominates, right? It's scattered all around the place, a sort of Where there is more sensitivity inside the institution, basically budget or interest, they're focusing there in a sense, right? Now, why does it matter? It matters because we're not talking about implementing one use case only.

We're talking about rebuilding the foundations with an architectural platform that allows to be AI plus. Starting from data and AI, basically to box out from the existing complexities. And intricacies and, boundaries of a business models for financial institutions, they leverage a FinTech in, in, in terms of evolving their their model.

So working on an individual use case [00:12:00] might make it hard to justify the investment required to really manage this technology and manage this technology means a governance, which is not just In principles, but it's grounded on, architectures and platforms, capabilities to interact with the problem.

So I think that on the one side, that all the banks are working on it, which is a very interesting, but on the other side, the majority is still thinking too tactically to really read the benefits for everyone of these new technology, let's see how these unfolds in 2024, we'll be curious to see at the end of the year, how many.

Modify their approach to become more enterprise wide and systemic. Yes.

Monica Millares: And then I think everyone, like you say, but this is me not based on research. My assumption is that people are experimenting, right? It's Oh, let's try here. Let's try there. But I think there's still not a [00:13:00] clear understanding across all teams, across all levels within the financial institutions on what AI can do for us.

as financial institutions. So can you expand on what are these multiple benefits or multiple use cases on how AI can help financial services?

Paolo Sironi: That's actually interesting because, AI is not a new technology. So generative AI only demonstrates the maturity of this technology. We've been also releasing a piece of research during World Economic Forum together with the research unit of Barclays Bank as well as Barclays and the City for Business Value of IBM paper, looking at the lag between when innovation and technology occurs and when the benefit is generated on productivity on the overall ecosystem.

So now this moment in time demonstrates that the technology is maturing. So it's not anymore the time for experimentation. Is the time for transformation [00:14:00] with this technology at core as the new steam engine put it this way. Now, there can be, of course, a variety of different use cases where financial institutions can generate value.

Talk more about financial institutions that cover a lot of fintech areas that individual fintech can see themselves inside of these these conversation. When we asked these 600 executives, all with responsibilities and decision making responsibilities with data AI strategy they would see the biggest benefits, 32 percent of them, which is the highest number said that risk and compliance, a financial institution is a risk management institution, right?

And compliance is typically a heavy burden. So in a sense, not surprised. And that is followed by client engagement. So the elements of communication with clients, which is also super important, then it development and the other support [00:15:00] functions. Now, that can be a ranking that's may correspond to everyone's perspective, or it's just the voice of the crowd.

Every institution will have a different point of view. And the reason is because it's not just about looking at value. Every value comes with its own complexity and its own risk. And complexity and risk can be very different according to the use case in terms of how much open it has to be, what type of data needs to be collected, how, if you like exposed you are with technology in front of final consumers.

So it's the combination of these three dimensions, potential value, complexities of organization, technology and risk. There is security, bias and hallucination that defines basically the tradeoff between these three elements. And therefore. Helps you understand what it takes effectively to implement on a foundational governance model the technology in order [00:16:00] to basically address a final user.

So that means the fact that the risk and compliance that corresponds to. On average, the highest potential value doesn't mean that is the easiest to be implemented, right? Sometimes higher value is more complexity. So all of these elements needs to be considered. And in this report, we tried to initiate that conversation with the readers.

Of course, it's just like preliminary elements because it's already 50 pages. You can type everything inside, but I think it can inspire people in the way they think pragmatically. About this type of innovation and transformation.

Monica Millares: Yes, I like that. Think pragmatically. So when it comes to compliance, can you give us examples of how we can.

Paolo Sironi: Okay. So let's say generative AI can do a variety of things, but the one that is more flashy that everybody sees is of course talking, [00:17:00] right? This lang is the large language model. So he's the. Make believe of human language, right? That is strikingly advanced that now compliance is based upon a huge corpus of information that needs to be read understood, reports that needs to be basically uploaded compared.

We already saw the capability of AI in terms of the more traditional models to work with document ingestion. Large language models and generative AI that is basically elevated significantly. And so you understand that the way you can interact with that complex domain now is much, much more advanced than what it has been in the past.

It's like the capability of the technology moved a bit at the time, then jumped up. Like this. So, in terms of [00:18:00] compliance, that's one of the typical cases, which is, basically understanding and optimizing the streamlining, the access to documentation, and the understanding the comparison upon all of these, then once you're inside that one.

You have other elements, which are, fraud anti money laundering, which are not just compliance is actually a foundational element of a financial institutions act. And that is where you see that generative AI can play both sides of the game. And on the one side, you can better understand what goes on in a complex system based off multiple communications.

But on the other side, Actors could also leverage technology for make believe, so to impersonate individuals and stuff. So there's a lot of catch up game here where you know that those people out there that have bad intentions tend to be more motivated to innovate fast compared to the others, which needs to be [00:19:00] more thoughtful.

So it is important that all of this is addressed. Now, because people are acting on both sides of the equation.

Monica Millares: Yes. That's yes. That is very true. We also need to think the bad guys are,

Paolo Sironi: One, one data point that amazes me watching the works during the world economic forum is in a panel, the CEO of JP Morgan Chase asset and wealth management, said that how many hacking attempts happen and go unsuccessful every day.

Attempts to break into the JPMorgan Chase system. You know what the number is, every day? Is it more than 1 million? Yes. Is it more than 100 million? Ooh. No. Is it more than 1 billion? She said 45 billion hacking attempts per day, and that's twice as much as the year before. Now, of course, that means there's a lot of [00:20:00] automation here happening, right?

And maybe not necessarily for private actors which might have a different perspective, not just an economic perspective. All of this is super important in our digital economy and therefore we know that some people might want to use technology in a bad way. We want to use technology in a good way also to preserve the value that we are creating.

Every banker must be an AI risk manager,

Monica Millares: period. Yes. And then this, makes me think everyone needs to think about regulation, right? Like we're regulated entities. So now. Of course, I'm assuming all the regulators will start asking different questions, e. g. How are you protecting yourselves about super automated attacks?

Now that you use this example, or maybe we're giving evidence for the compliance report, but it was AI generated and they are like, how do I trust that your AI [00:21:00] generated system is doing a good job? Basically, how did you think that conversation with regulators will start changing?

Paolo Sironi: That's always a dynamic conversation in this respect, I believe that the industry has the responsibility to be at the forefront of that conversation, right?

And so basically, not just to leave it, but to start it because it's important for everyone happened. Also, when a risk management made the first steps forward, regulation without a clear cooperation among all the parties can become a bit hard to. Okay, neurological. And it's not the fault regulators process in terms of complex domain and I think the industry has a responsibility to make sure that in an open discussion, all parties basically find the, element in common.

And the element in common is to make sure that a good work is done for the value of consumers in a secured way. And that's it, because that pays out anyway, the long term for everyone.

Monica Millares: Yes. I'm talking about [00:22:00] the final consumers. I think all us. FinTechs or as financial services, it's our responsibility.

That's why we exist, right? To help people have better lives with money. Part of that is accessibility. And part accessibility can be defined in many ways. One of them is make it easy for everyone to use. whatever we're using, whichever service it is that we're offering.

Paolo Sironi: Make it easy. It can be a delicate, expression.

So make it understandable. Okay. So that's why it has to be transparent. Sometimes a friction is needed to allow people to think. Because if you think about elements lab and operator, maybe it goes too fast. Okay. Sure. So the point, however, is that what is that friction about? What is that moment of reflection?

What type of value it creates as long as it is transparent and allows people to understand more what they're doing, that means that we're all going in to the right direction. [00:23:00] Okay. So it's a collective effort for for everyone. But then at the same time, sometimes it's too hard for people to do that right there in simple terms.

So, that's when technology allows you to access concepts, information, and, ways of looking at a problem that otherwise you will be precluded from, right? So it puts you more in touch with the problem that you have to resolve.

Monica Millares: So just to build on that, how could then AI help us improve the customer experience such that it has the right level of friction?

Paolo Sironi: Okay. So let's start from a more abstract, if you like. So without getting into the details, to by representing how banks have been investing in technology. Okay. To address a digital experience in the last 10, 15 years, let's say that 10, 15 years ago, all banks started discussing the fact that they needed to be [00:24:00] more proactive online.

And definitely as the smartphone came to life to be on mobile, so mobile technology is a digital technology by definition. I think the first thing that most banks have discussed was how can I power up their mobile app in a way that it works fine. So that became a primarily a shift to cloud. Okay. So you need to use cloud because now you're operating with a mobile apps in very different ways instead of allowing for access to on prem technology.

So it's a cloud discussion at the very beginning. Banks were very. A concern about cloud that they learned also regulates that I can do that. So they accelerated, but just lifting and shifting a banking experience on cloud. That doesn't mean that you're delivering a good digital experience. And that's when, the understanding that the data was relevant to make it more personalized.

Became an everyday discussion. So that is when artificial [00:25:00] intelligence, machine learning, sometimes deep learning, natural language programs. So, that type of more traditional AI, if I can call it this way came to the forefront of investments and discussions. However there is still a complexity here or a delicacy.

Not only the net promoter score of banks has really improved significantly. And it's it's concerning. It's not top of of the industries. But the majority of people have a harder time of interacting for the most important decisions with a digital medium. As I mentioned before, let me reiterate it.

What is digital technology? My iPhone or your Samsung. This is a technology of the demand. People don't go on Amazon to see what happens. They go there because they want to buy one of my books. And you can also likely find that for free. This report, there should be an Amazon link. But the most important actions, which correspond to the [00:26:00] most important revenues of financial institutions operate in an offer.

Oriented framework and an economy where banks are pushed to clients insurance push to their, clients financial products and their capability is some, somehow to overwhelm me because people never have the time and understanding, right? So now if you want to elevate people to use digital technology with a good experience to be more autonomous, what you need to do is to create better element of communication.

And that's when you see that now we go from cloud through traditional AI into. Generative AI, because Generative AI is about communication. Of course, we need to make sure that every communication element is framed within clear transparency and compliance framework in a way that the people are not, if you like, hallucinated or, that you don't generate if you're missing out. That's always important. But in any case, that was the missing point. And then now we start to [00:27:00] see that it is coming to life. The possibility to personalize a true communication, the decision making necessity of people when they interact with a mobile app, and they don't necessarily have another type of interaction with a banker.

Still some work to be done, but for the first time we see that is basically conceivable.

Monica Millares: Okay, so I have a very basic question. As I was reading through the report, there's a section that you talk about communication as a service, and you also make reference to decision making. And I was like, Oh, I'm a bit not confused, but I was like, Oh, it's the very first time that I hear the term communication as a service, especially within financial services, so if I paraphrase, let's see if I understood correctly, if I paraphrase what you just said, it's basically.

Communication as a service allows us to use all the information that we have about customers, all the capabilities that we have, such that we create better. [00:28:00] understanding from a customer perspective on what are the interactions that they are having with us as financial institutions and

Paolo Sironi: with their money.

I would say on both sides. So the researchers of the European Central Bank a few years ago wrote a paper titled financial intermediation with technology, but so that what's new that also leveraged in my bank, safe and tech on platform economies to create the bank and invention quadrant axis.

What they said is that financial institutions exert market power when they excel in information and communication, which are the two axes of the product. So what's information, what's communication information is about all the data. About a client that typically reside in the core banking of an institution, you can think about the adverse selection, so basically identifying who is worth it for credit and for how much, or information about payments and transactions around payments that basically identify elements of knowledge around the [00:29:00] customer.

And if you think about in particular, the European banks, but you can expand across the world as this is a mega trend. But in, in, Europe and in Japan, banks had a harder time in generating business value through information, the core banking elements, because of very low interest rates, very high credit risks, amount of interest will go up, the credit risk increases, so it's more complex.

So this is when the European central bank researcher said that there is a shift that is clear in Europe that moves from information to communication. Communication is people talking to people. It typically goes around that more complex decisions for your loans, mortgages, or investment products for retirement, for opportunities or insurance can be like insurance or something else.

They typically are very asymmetrical because the domain of expertise is much larger and deeper. And it requires more if you like thinking and that typically is done people with people, right? So now [00:30:00] it is difficult to scale up that communication element. And how do you basically allow all of your network of employees working in a bank to be elevated enough in order to make those conversations in a way that are effectively impactful and get it through.

So now. The community, the, if you like, systematization of communication has always been complex because it's human related, but now that can be elevated. And I'm not saying really, it has to be substituted because there's a lot of these cases where employees that can be helped basically to have a better conversation with the customers because they are elevated to understand faster to master the domain of expertise.

And that's where communication as a service becomes important. If you like, it is where the bank also addresses the bulk of the most complex decision making that can happen in society at the micro or the macro level.

Monica Millares: Okay. Can you, sorry, I followed most of the conversation except the last sentence when [00:31:00] it comes to decision making.

Can you re explain slash paraphrase? The bit of how communication as a service helps with decision making.

Paolo Sironi: Okay. Let me tell you, the biggest problem that many FinTech had is that they misunderstood the capability of people to self direct themselves. So they misunderstood the role of human relationships.

And I can do this in a very simple way. Looking at the, main business model of a bank, there are four areas, payment, credit, investing, and insurance. Mention them in order of a simulator of information. So the complexity of the decision making payment is very simple. You click and buy your Gucci bag.

So as long as it is convenient or fast, you do it. That's why the biggest unicorn in fintech started in payment is the pay tech, right? It's a volume business. RJ should be an uber volume business. Take it. The second next to this one, which is still fairly symmetrical is credit. And it is fairly symmetrical because you get money, right?

And so if I tell you, [00:32:00] Monica. Go to my IBM. com slash IBD where I host my IBM papers, click on the first page, and I will give you 1, 000 free of fees zero interest rates. You give back the money to me whenever you want you may do it. Many people might do it, but maybe some of my IBM colleagues as well.

And I'm not sure about the credit worthiness, right? It becomes a big mismanagement problem, unsecured lending, buying operator Goldman Sachs with Marcos that made the mistake. It's not about giving away money, right? But it's giving away money in a way that is good for people. You understand the risk about that because some people will have problems in, the end, but it becomes a bit more complex, but still people.

Can go on it. Credit tech was the next to follow because people get the money. But when it comes to investing, it's a very different problem. Let me say, go back to the website, click on it, and you will give me 1000 of your dollars. And I will invest that into a model portfolio made of cryptocurrencies, European stocks, and Chinese stocks.

Now, many people in [00:33:00] the audience will ask, is it true that Paolo worked in quantitative finance? Did he write books on portfolio theory? So there is a gap, right? People need to trust them. They may not have, they may have a hard time, the majority of people to understand how to make the decision. So they need to trust the conversation.

They need to trust people when they made the decision. And that's the reason why typically that is not a mobile thing. The execution maybe. But the decision making process requires trust among humans. And I give you the easiest example, which is insurance. The most asymmetrical of all. Because it's not even about risk, it's really about uncertainty.

Think about life insurance. Monica, I want to sell to you a life insurance contract, so why did you tell her that you're going to die? Now I'm sorry for that. And I guess you might not want to have that conversation with LGBT, you prefer to have it with me. So that means that when we get deeper into you're facing uncertainty, which is the core of banking that is on communication.

So relationships can be more and more important now, even though communication can now [00:34:00] start working on sorry, technology. I guess I'm working on communications. Still, the human element is foundational. For that for bridging that trust gap, but many more human conversations can be elevated, right?

In a way that they are accessible for for people. And that is really the hard part to be digitalized. And that's the reason why it is not so easy to digitalize investing in insurance compared to. Payment and somehow credit past, if you like auto insurance, if it's compulsory it's compulsory still, 90%, if I'm not mistaken, of Italians still prefer to talk to a broker.

And in our research, we identified the reason why people acted this way only in the UK is the most advanced market in terms of mobile, access 60 percent if I'm not mistaken of a consumer's buyout insurance, which is compulsory, more or less early. [00:35:00] online, instead of talking to a broker, but still a large percent of them talk to people.

There are elements of some of the reasons that reveal those in these regenerate banking with AI paper, and they're worth understanding because that's where the next frontier lays in terms of digitalizing financial services communication.

Monica Millares: Thank you. Thank you for expanding because now I like.

Now I understand, thank you. And I'm sure listeners will understand as well. So I want to change the conversation a little bit and move away from customers and tech and move back to people like the people working in financial services as such. So there's this conversation of fear and opportunity at the same time.

Fear that you will lose your job and opportunity that will make. Make you a superhuman. So what we could have the debate, like I am more of the, it'll make you a superhuman. You have to embrace it. That's it. But you talk about re imagining the workforce experience as such, [00:36:00] boosting productivity and rebalancing costs.

What does that mean?

Paolo Sironi: First of all, of course technology, like every technology intersects, the way we work and in some cases transforms it to basically replace it, right? So things that we were doing before I, by hand that we, now do in a different way, but I do believe in the majority of cases that It is not that AI will replace bankers.

The key point is that bankers who can use AI will advance more and faster than bankers who cannot use AI. The same now everybody asks you whatever level of, job you want to have if you know how to use Excel, right? So some elements becomes, if you like in and out. Sorry. Learning how to use AI makes a [00:37:00] difference among people and people that don't know how to use AI basically will be recruited from opportunities.

When we look at the economic model of banks, we looked at that globally, okay, all banks with more than 50 billions of assets of total assets, the operating expenses have been growing over time. There's also an inflationary element, but effectively they've been spending more. But the portion of money which is utilized for employees compared to the one investor in technology grew as a percentage of this basket more than the technology one.

And I think it's important because it's about people, but That creates always a tension in institutions because how much can you turn efficiency into cost reduction as the cost income is really, unsustainable for, many banks out there. And so the point here is where can technology allow people to contribute more and better?

To the success [00:38:00] of the institution also as an element of self fulfillment. So all of those elements are important that are at play when it comes to unlocking the value through technology, technologies such as Generative AI. Now, there is a very interesting data point inside that is a global outlook.

That I wanted to reveal here in this podcast, then we've been asking us the IBM Institute for business value 3000 executives every year since 2016 about them, which are the skills which matter the most for the workforce. And we identified 14 of them. And in 2016, the top two were STEM, mathematics, engineering.

High level computer science and basic skills. And then we're followed by down creativity and other elements like team workability stuff. We asked the same question in 2023. And where STEM skills [00:39:00] rank. 14. There's always a wow, right? When you say that now and people think that, oh an engineer is not creative.

So creativity goes up and creativity is still there, mid of the the, ranking would went up. Is the ability to prioritize the resolve complexity and teamwork ability. Why is that? This is not, to be seen as a static picture is, a flux right? It's not that stem is not important, but I think that the industries are realizing that if they don't resolve the complexities of the way they work before adding new talent with the move the needle, actually this talent will not, will be precluded from contributing.

Many times people say, I'm frustrated. There was hired I had this ambition, but then I cannot work. And the same goes with advanced technologies. If you don't resolve the complexities of the operating model, you can plug in more technology, but you may not resolve the, actually, you can even reach an event horizon.

We're saying in this paper very fast [00:40:00] you add capabilities to develop on complex systems that the last causality. From their origins at the original business purpose, what happens? You can even go faster in building complexity. So it is super important that the operating model discussion is at core of this transformational moment, because only by addressing those the rest can go, that's why I think that there is a higher request for understanding how to prioritize how to work in teams how to.

Plug in those capabilities, they start transforming the core of the complexity of the operating model so that the rest can effectively add value for the good or for the bad, everything is on the table. Okay, the old moving parts, so nothing is is left behind. Everything needs to move at the same time.

Monica Millares: Yes, and I think that's something that we all need to be comfortable with uncertainty and change because. [00:41:00] It's just accelerating the pace of, uncertainty and change is just, it used to be just for FinTechs early stage. Now it's everywhere.

Paolo Sironi: Yes. The global outlook 2023 was uncertainty is the only certainty in essence, right?

Also on the global stage, what matters is to be transparent as much as possible because transparency beats trust, right? You will not resolve uncertainty. So there's always. There will always be question mark open, but when you know, transparently, what is the effort, what everybody wants to do, so then you're more inclined to participate, right?

And to contribute, otherwise you feel left out. And that's not good. Nobody should be left out. No,

Monica Millares: exactly. So I'm very conscious of time. This is a fascinating conversation. Probably it should be like a three hour podcast. But then if we think about, you used the word pragmatic at some point in the conversation.

If we are like, okay. I listened to you guys, [00:42:00] you've been talking for almost an hour. Can you give me a summary? What are the steps? How can I be pragmatic? And what can I take back to work when it comes to thinking about implementing AI? How do I exactly simplify the complexity work with the team? How

Paolo Sironi: do we do that?

Okay. So first of all, we said that every banker must be an AI risk manager, right? And that's is needed because there is management, the government is working to end. There are four principles that matters. First of all bankers have to manage for value more than ever. They ask themselves where value can be generated by the application of technology.

Then they have to manage for complexity of the innovation. Sometimes it's not feasible or it's very complex to implement it for a few reasons. Technology, maybe you don't have a data access that matters. Maybe you cannot create, transparency across the algorithms. Maybe you have organizational [00:43:00] complexity.

So you have to change the way people work and you're not ready for doing that. And then you have to manage for risk because cybersecurity first we are expanding the vulnerability space. So we need to make sure that we also plug in the capabilities to make everything more securely. And on the other side, there is always a risk of bias and hallucination, especially when technology is used and it can go viral.

So they need to be considered first. So managing for value, for complexity, for risk enables to do the fourth. To manage for scale, because as we said, this is not just a use case by use case is about AI plus. So putting AI and data at core to change the way we work to unlock everyone's potential talent, employees clients to participate into the banking journey.

So we crafted a 10, Guiding actions if you like chart that basically helps to visualize and conceive what has to be done [00:44:00] every year is like a recursive process and looking at AI, some example in general, but it's everywhere. Because first of all, there is always an element of exploration to understand that the priorities that a bank has, again, is not a technology in isolation.

First of all, which are the priorities of the bank and how AI intersects those priorities then is about how to integrate at core. Data and AI inside the operations is not a use case in isolation. It can start with, but you need to understand quickly how basically you can work on it systematically by allowing some space for maneuvering areas, which are a bit more on the frontier.

There is always something which is a bit more on the frontier, right? That they cannot fit the process of standardization. And so that's where effectively you learn by making, and you need to create a feedback that enables you to quickly understand how you can further [00:45:00] improve your governance and architecture.

And then it's when you start scaling with AI, basically get into an AI plus framework. So it becomes a scale up approach where you are building in progress your AI factory. That means you elevate every colleague to understand AI, the risk management of AI, and how to risk manage with AI, and then you ground it on a platform that enables you to further contribute and to restart the process to make it better.

Better is a continuous process of building culture. It's just that the operational culture now starts with AI and not is not executed with AI only.

Monica Millares: Yes. Thank you so much for explaining basically the book in plain English and words. I think it makes a great compliment. You should even think about having the audio book, not like reading the book, but explaining it in your own words because it brings it to life.[00:46:00]

Yes, exactly, because it brings it to life in such a different way. It's Oh, I read it, but it's complex. It's new concepts. It's new language. But then speaking with you, it's Oh, now I understand. Better. So I love, this conversation. Thank you.

Paolo Sironi: So I'm grateful for this opportunity.

Everyone can download. A free copy of our research is on the ibm.com/ibb, and then there will be the banking and financial market pages. Hope you can also add the link, please, podcast. And I invite everyone to listen to this podcast. Monica's podcast are the greatest and latest in terms of understanding technology and business transformation.

So, happy that I can. Do this reading together here with you in your audience.

Monica Millares: Thank you. And I'm even actually, I am creative. And I like studying. So I'm even thinking Hey, I'll create a mini study group so that we go through, we read the book, we go through the podcast and then we

Paolo Sironi: discuss.

Monica, that [00:47:00] word you use resonates to me. People don't have time to read anymore. Even platforms like LinkedIn became very short posts, not articles anymore. In the media, you've got the image, everything shrinks to the limit of a tweet, but reading is important. And you know what? Even people that read and not necessarily study, but it is about studying as well.

Dividing element and element, getting a deeper understanding. That's why I always like to share printed copies of the material, because I think that, again, the reading on paper slows down and we need to find time to study. That means slowing down our interaction with the problem. It is super important. I think a bit of a friction slowing down our reading time adds value because we all decide to study, don't simply read.

Yes. No,

Monica Millares: totally agree. That's what I've been doing lately. But it, my previous book, it's I bought the [00:48:00] book, highlighted the book, and now I'm taking notes about the book. I'm studying the book. So I think I'll do the same with your book. It's been an absolute pleasure coming in the show, Pablo. One very last question.

This is the universal question that I ask at the end of this. podcast that by now it's becoming a tradition. So if you could change one thing in fintech to make fintech better and have positive to customers, staff, and investors, what could that be?

Paolo Sironi: Don't think as a startup, think as a scale up

because this is not about, optimizing venture capital money. And I know it's difficult because you need money. So typically I gave a check of the money. This is about contributing to the transformation of banking, which is a core industry that goes across every other element in our society. If we can make banking better, I'm sure that we [00:49:00] can improve the quality of all of our other economic ecosystems.

But that requires a scale up mentality, not just a startup mentality.

Monica Millares: Definitely. I love it. I also think about that all the time, that it's money touches every single, industry. That's why it's so important. So that's why we have people passionate about it, like you and I and many others.

It's like we geek out. It's been an absolute pleasure. Thank you, Paolo. I'll put all the details you referred to your books to other research papers. All of that will go in the show notes, the transcripts so that people can study as well. Like, all of that will be included. Thank you so much, everyone. Thank you.