Mike Walsh: GenAI Will Unlock the Potential of Human Intelligence

Jim Marous (00:00):

Welcome to Banking Transformed, the top podcast in retail banking. I'm your host, Jim Marous. I am thrilled to have global futurist Mike Walsh, back on the Banking Transformed Podcast. Mike's expertise in AI powered organizations and his vision for the Fifth Industrial Revolution offers invaluable insights for businesses navigating this rapidly evolving landscape of technology and knowledge.

Jim Marous (00:39):

In this episode, we explore how generative AI is reshaping the future of work, the delicate balance between humans and machines, and the transformation potential of AI as the foundation for modern businesses through cognitive learning.

Jim Marous (00:56):

The future is being rewritten by AI, but still led by the creativity of visionary and adaptive talent. The lessons learned today apply across all sectors. Obsolete skills matter less than learning new ones.

Jim Marous (01:14):

In our discussion today, Mike Walsh shares insights on how businesses should adapt to the AI centric environment stressing the importance of a radical cultural shift. The discussion will also cover the changing nature of employment and the essential skills required in an AI dominated future emphasizing the need for continuous learning and adaptability.

Mike Walsh (01:36):

So, Mike, you've joined us every year since our podcast started discussing how data, insights and AI are changing businesses. Well, in November of a year ago, ChatGPT was unveiled bringing to life much of what you had discussed through the years. How do you envision generative AI transforming the future of business and its operations beyond what was possible only 12 months ago?

Mike Walsh (02:03):

First of all, it's wonderful to be back on the show again, and I'm very honored to have been such a repeat guest. I have to say, though, I have mixed feelings about the fact that AI has become so prevalent now. It was a lot easier and maybe more interesting talking about it when less people were interested.

Mike Walsh (02:18):

But now I think, as you mentioned, because of ChatGPT, it's had such an extraordinary take up. Everyone not only feels that they understand what AI is, I think quite a significant majority feel like they're also experts in it, which is quite dangerous.

Mike Walsh (02:35):

And I think we're really just at the very beginnings of trying to understand what an AI powered revolution might look like when it comes to transforming organizations. And the reason for that is that for many of us, generative AI has become a tool for writing breakup emails, rhyming birthday greetings, or summarizing memos you can't be bothered reading.

Mike Walsh (03:02):

Most people have really just begun to scratch the surface in terms of what it really can be used for in terms of not just summarizing information in new and interesting ways, but really becoming an indispensable platform for generating breakthrough insights.

Jim Marous (03:18):

It's interesting because in a lot of your discussions on your website and articles you've written, you really talk about the ability for generative AI to build better personalization, build better dialogue, and actually in a way, humanize things as opposed to making them simply more efficient. How do you see AI progressing, expanding the human capability and cognition rather than just improving efficiencies?

Mike Walsh (03:47):

Well, there's many parts to it, and I think this is a fantastic question and way of framing it because one of the biggest traps of any form of technology, I think, is to take what we do today and try and do it 10% better. That's the easiest business case to always make if you're trying to justify some new spend.

Mike Walsh (04:05):

But it is the most dangerous thing to do when it comes to a kind of a fundamentally disruptive paradigm transforming technology like artificial intelligence. AI, I think today is already giving us the means to not just speed up our existing workflows, but to really reframe the way we think. And I call this metacognition, thinking about thinking.

Mike Walsh (04:34):

And in our previous discussion, we were talking about how we both use generative AI tools like Claude and ChatGPT. The way to use them is not to generate content, but to actually become a sparring partner, become a cognitive assistant to actually challenge the work you've done, to get it to read it, to give you a kind of some constructive feedback to kind of push your writing, your thinking into new directions.

Mike Walsh (04:59):

If you're just asking it to write a kind of six paragraph essay on a particular subject or an article, and that becomes your product, then really what you're doing is just taking the average result of a particular thinking in an area and trying to submit it as your own.

Mike Walsh (05:15):

That is not breakthrough thinking, and it's certainly not improving productivity. Where I think things are going though next is going to be much more interesting. And that is really thinking about AI as not just an extension of ourselves, but as an autonomous agent who can undertake complex planning, thinking tasks in the real and virtual world.

Mike Walsh (05:39):

And when we start seeing generative AI in that form, that's when things will start to really look quite different from the way they are today.

Jim Marous (05:49):

It's interesting, I was having a conversation, jeez, just about six months ago with Brian Roemmele, and we got into this discussion about could generative AI actually be your individual agent?

Jim Marous (05:59):

And we're taking banking as an example and saying through the questions that the tool can ask, through capturing the answers that it gets through transactions, behavioral trends, things from outside and inside the financial institution. Could this become your agent at the financial institution that could build better go forward solutions on a personalized basis?

Jim Marous (06:19):

And in other words, if done well and if done with continuous interaction with a client, could it actually increase trust as it remembers things you said. So, it does what humans are meant to do and can still do in conjunction with the tool. But can it actually build separate units, separate agents in each transaction for each customer build in a way that is simplest? Do you see this as a possibility?

Mike Walsh (06:50):

No, absolutely. In fact, I would go further and say it would be insane if it doesn't. But I think the key to understanding this is that it's not just a bank that will have a virtual workforce of agents that are in a sense, an extension of their call center or their human agents, individual consumers and professionals will have their own digital twins, their own digital doppelgangers.

Mike Walsh (07:13):

And this is where you really get the hyper personalization. You're not relying on the bank or the financial institution to have that. You're actually relying on your own highly trained highly autonomous personal digital extension of yourself. And this agent will absolutely be reading your emails, be writing your correspondence that isn't mission critical will be applying for bank loans, credit cards, will even maybe go on job interviews.

Mike Walsh (07:45):

It'll essentially be your consiglieri in the digital world. And because of that we are going to see consumers trusting their data to be used much more because they'll actually have control over this agent. And if you want to see like a preview of what that's going to look like one of the most interesting devices to come out of CES was the Rabbit R1.

Mike Walsh (08:09):

I'm not sure if you came across it kind of looks like — it was incredibly designed. It's like this little red thing. It's kind of like a phone that isn't a phone. And the idea was, is that they wanted to build a device that was natural born in the AI era.

Mike Walsh (08:24):

Rather than a large language model, they call it a large action model, because what the rabbit does is it actually learns how you use particular apps, whether it's calling an Uber or ordering food, maybe in the future opening up a bank account and it has a little camera that watches how you do it. And then it actually acts like you on those apps through kind of a secure tunnel in the cloud.

Mike Walsh (08:49):

So, this is sort of a personal version of robotic process automation that we saw in the enterprise for years. But what it allows you to do without these app developers creating sort of a whole complex LLM model themselves or building their own APIs, it really allows you immediately to kind of create a digital twin of yourself that's acting on anything that can be done on an app today.

Mike Walsh (09:17):

So, this is the beginning of a world, I think, where we will start to see AI acting as this sort of digital personal agent.

Jim Marous (09:26):

It's interesting because again, when we were talking about it in the banking order in the world you just mentioned, it's also the questions that are asked. The testing that goes on automatically. The first tests of ChatGPT and other generative AI tools, you realize that it's not simply asking a simple question and getting a really complex answer.

Jim Marous (09:48):

The more specific, the more unique your question becomes, the better the tool becomes. And it's interesting because in your world also where you have this digital twin, you can actually set up different digital twins with different identity authorizations as well where I want this agent to know everything and share everything with everybody. In this case, I want my identity not to be as full, not to be as robust, but only as robust I need for this purpose.

Jim Marous (10:19):

And that's where I think the trust issue is so big in the whole generative AI and in AI tools where organizations have a question about it, governments have a question about it, humans have a question about it, but it's really about that learning process that that is the key to that trust element.

Jim Marous (10:38):

And it also, what I'm wondering, because I think you mentioned this in one of your recent articles, it can change the value proposition from simply monetary exchange to having values for risk, having values for how much it knows, having value for how quickly you can get to the answer base. It really changes everything.

Mike Walsh (10:58):

I mean, I think people tend to anthropomorphize AI and this is where everything gets confusing. It isn't so complex to imagine an agent only be able to speak about certain things to certain types of digital entities. It sends the idea of having Chinese walls in memory.

Mike Walsh (11:16):

So, the things it says to a potential Tinder date, where the AIs are talking to each other, it is very different to when they're talking to a financial institution, trying to prove credit worthiness. Not that the bank wouldn't be fascinated in some of your Tinder discussion, because it could probably build a more complex model of your risk behavior using that information.

Mike Walsh (11:39):

But it would potentially break a whole bunch of credit reporting rules, and it becomes even more important with medical data. Anything that's HIPAA compliant. So, I think compartmentalization of data is going to become very normal, very quickly.

Mike Walsh (11:58):

But I think when it comes to AI regulation and this is the point you were referencing, these virtual agents, let's be very clear, are highly unlikely to be self-aware in any respect. In fact, I would argue although they could do a very good job of responding like us, they will not be us.

Mike Walsh (12:21):

And if under any circumstances it started to appear that these things were in any way self-aware, I would say we'd have a moral duty to actually delete because frankly, they would be non-useful at the point where you actually need a psychologist to get these things to work properly. So, for me, the non-cognitive awareness of these agents is what makes them useful. We don't want digital versions of ourselves that are in any way self-aware.

Jim Marous (12:53):

So, that's interesting. When you look at organizations today, you have some that are buying all into AI tools and ChatGPT and generative AI. Then you have those that are the exact opposite extreme where they're not buying in until it's all proven, which is interesting concept.

Jim Marous (13:13):

What is the difference, and how do you see the marketplace looking right now with regard to building from an AI foundation and for those organizations that can actually maybe start reorganize the organization around AI, what kind of advantages they have versus those that are waiting for the test to be tested?

Mike Walsh (13:34):

Well, Jim, I think you got to take a step backwards and go, what is the point of being an AI powered organization? And for me, the only point is not to make Microsoft shareholders any richer, the point isn't to try and consume more of the coolest product, and I think that the benefits of putting out a press release to say that you've got a generative AI strategy that kind of expired nine months ago.

Mike Walsh (13:57):

So, just being the latest person to do something with generative AI is a bit like doing something with the metaverse a year ago. You got a little bit of kind of kudos for it. Now you look like a moron. So, I think the only reason that you would embrace AI is because you've got a coherent plan for increasing your enterprise value.

Mike Walsh (14:19):

So, my theory is that in the next 10 years, there's going to be a very stark definition in terms of winners and losers. And in the old days, the distinction was between technology companies and traditional companies. So, a traditional company-

Jim Marous (14:35):

Traditional world. Yeah.

Mike Walsh (14:36):

Yeah, I mean, but you'd value a traditional company, whether it was in agriculture or logistics or manufacturing or consulting on some kind of reasonable multiple of EBIDTA. It wasn't rocket science and you could kind of haggle over that, but it was kind of a fairly established number.

Mike Walsh (14:55):

In the digital world, who knows what these numbers were? I mean, some of these digital, even now, if you look at some of these AI valuations, I feel like they're just multiples of how many Xboxes that the kids have got in the rec room, they're just making up the numbers.

Mike Walsh (15:10):

But I think this is going to shake out in the next few years. And you are going to have, in even very traditional industries, the difference between companies that are fully AI leveraged and those that haven't.

Mike Walsh (15:22):

So, there's going to be an AI leveraged player in agriculture that's going to have a valuation that might be similar to Tesla. You can have something similar in life sciences. You're can have something similar in logistics and financial services, just like Tesla is an outrageous valuation for an automotive player.

Mike Walsh (15:39):

But it's because people said this is a AI powered version of an automotive player. Although, to the degree to which Elon Musk plays along with that may depend on whether they give him more stock or not.

Mike Walsh (15:53):

So, this is the kind of dynamic I think that's going to play out in every industry. So, if that's the game that is worth leaders playing for, then the next question is, well, what makes a company truly AI powered? And for me, it comes down to four factors, which is scale, speed, sustainability, and scope.

Mike Walsh (16:13):

And these are the kind of areas where these AI organizations, regardless of their sector or field or market, are going to seek to differentiate themselves.

Jim Marous (16:26):

So, with those components of a successful AI business strategy, how does an organization who’s I'm going to say legacy, and we can define any way we want, where do they go upon moving that step? Because those four, while there's four components that are all understandable, there also can be overwhelming seeing where we are today. Where does our organization start, and how do they succeed in the short-term as they move toward a larger perspective going forward?

Mike Walsh (16:58):

Well, I think the question, the kind of the starting question that any leader in any organization, regardless of industry has to ask themselves is whatever investment we're making in AI that we're planning, how are we changing the system of work? And when I say system of work, I mean the kind of the process by which we create and deliver value.

Mike Walsh (17:22):

If we're just kind of deploying a chat bot for customers to screw around with, or for people to ask interesting questions and generate reports, is this actually changing fundamentally the way that we design and deliver value?

Mike Walsh (17:35):

And if it isn't, then honestly, you are unlikely to move the needle on valuation. And in previous industrial revolutions there’s always been the winners and losers in the same way that the people that took that disruptive innovation, whether it was steam or electricity or computation, and were able to actually not just do what they did before a bit better, but to actually do something new and different to change the underlying economics.

Mike Walsh (18:09):

This is what Henry Ford did if you look at the Highland Park facility, it wasn't that he just used electricity and kind of electrified his manufacturing process. He goes, "Well, actually, how can we redesign the manufacturing process? What should a manufacturing plant look like now that we have electricity? Can it be more decentralized, more distributed, more agile?"

Mike Walsh (18:30):

And these are very 2023 kind of phrases, but this is his genius. Because if you go back into the kind of 1930s, this is how he was thinking in very 21st century ways of how this disruptive technology changes the way we work.

Jim Marous (18:47):

So, when you say that, then it really gets down to winning organizations will be those that can reinvent what their end product is, but in a brand-new way. Because as you said, simply electrifying what you had in the past just gets you more faster, doesn't necessarily make it better. So, it's going to take leadership and culture to really rethink what you're delivering and how you're delivering that. And that really takes complete relearning internally, doesn't it?

Mike Walsh (19:20):

It does. I mean, because you're really changing two things at the same time, which is difficult. It's a bit like patting your head and rubbing your stomach. Because you're not only changing what you make really turning your product into a platform, essentially, you are changing how you make it as well. And really you need to be able to do both.

Mike Walsh (19:42):

Sothis is really, I think, difficult for people to conceptualize. And when people do it, people use that as a metaphor. I mean, if you go back, to the early origins of Uber, people will ask themselves, "Well, how do we become the Uber of X?" How do we do what Uber did? But let's do it in a completely unrelated field.

Mike Walsh (20:01):

And they kind of missed the point. The point was not that Uber was a template, the point was the process by which Uber became Uber. Someone said, okay, there's a new emerging market for this, which means we can use technology and computation in the internet, smartphones to basically do something in a completely new way.

Mike Walsh (20:20):

But they didn't just do that. They changed the way that the organization was set up, the way they used data, the way they scaled, how they moneyballed regulation. I mean, not all of it was completely ethical, and it didn't always make it a great place to work, but essentially, they were doing those two things, reinventing the product to the end consumer, and changing the design of the organization and how it interacted with the ecosystem.

Jim Marous (20:44):

So, when you look at that and you look at where we are today, what are the biggest challenges that you're seeing as organizations try to transform from simply being digital and being, let's say fifth generation aware, the personalization, the engagement, the interaction. What's going to be the biggest challenge that gets in the way of that transformation?

Mike Walsh (21:11):

Look, the biggest thing everyone's facing now is scaling up. I mean, I think if you weren't doing some kind of generative AI pilot in the last 12 months, you either were living under a rock or you're being deliberately obstinate.

Mike Walsh (21:27):

And you're right, there is a whole bunch of organizations who are quite rightly terrified of what's going on. And there's a bunch of leaders who are hoping they can retire before it's on their plate. And for those people, let's face it, when this thing becomes something that is as prevalent and as acceptable as social media, then it's probably almost certainly already too late.

Mike Walsh (21:50):

So, you should have been using the last 12 months to basically be flexing your experimental muscles to kind of learn what you don't know, and to try and understand what your organization's absorptive capacity is.

Mike Walsh (22:04):

And what I mean, absorptive capacity. What is the kind of the flow rate by which new ideas can penetrate your organization and new ways of learning. Hopefully, you've figured out what that capacity looks like in the last 12 months.

Mike Walsh (22:20):

Now 2024 and beyond, you should be thinking about scaling that up. How do we actually take those experiments, those pilots, those projects and actually do something which really transforms an end-to-end workflow that changes the design of an entire division that is the basis or the kind of the fundamental technology behind a whole new customer facing platform, that should be your play for this year.

Jim Marous (22:50):

So, it's interesting, when we went to the digital phase, and we're still nowhere near done with that, but when we went through that, there was a real change in the marketplace where the consumer really got control of what they wanted and had examples out there in the marketplace that said, "Why can't you do it like X?" Whatever that X company was.

Jim Marous (23:11):

And we see organizations saying those same questions, but the consumer's really been driving their expectation level. And COVID amped that up. But when you're getting to this next phase, where will the consumer feel this first and put pressure on every organization to perform in a way that's similar to the leaders?

Mike Walsh (23:33):

It's a great question. And you kind of can think of this broadly as the consumerization of experiences. And this has been a journey I think we've been on for at least 30 years. Consumers discover smartphones and they wonder why the corporates give them these ridiculous devices that they can't do what they can do at home.

Mike Walsh (23:57):

And then we have bring your own device, and it kind of transforms enterprise it consumers discover the amazing world of cloud-based entertainment apps like Spotify and Netflix, and they wonder why is not all experiences like this. Whether it's banking or insurance or transportation.

Mike Walsh (24:17):

So, in these kind of cutthroat, dynamic consumer markets where experience is everything and personalization drives profits you win or lose depending on your ability to deliver these world class experiences.

Mike Walsh (24:34):

Often the enterprise world where things are sold more slowly and there's lock in and kind of an inertia and a willful blindness to the end consumer or customer, it can take longer for these things to percolate. So, that's been at least 30 or 40 years, we've seen that dynamic in this new AI revolution.

Mike Walsh (24:58):

What we're already seeing is people realizing the power of being able to ask a question and get an intelligent answer. And so, this changed overnight people's expectations of now dealing with contact centers with corporate chat bots. It's becoming obvious that you are talking to something highly intelligent like ChatGPT or Claude, and then you go to a standard chat bot and it's like talking to someone with a with-

Jim Marous (25:27):

Voice prompts. Yeah.

Mike Walsh (25:28):

Yeah. It's like talking to a digital entity with a lobotomy.

Jim Marous (25:35):

Something to think about, yes.

Mike Walsh (25:36):

Who's not only inanely stupid but seems intent on directing you down a whole rabbit hole of banal irrelevance. This thing is designed to actually make it impossible for you to answer your question or complete your task. It is designed as a barrier, for you to get something resolved.

Mike Walsh (25:59):

So, I think this is where the kind of the next level of where customer expectations are sitting, and it's already a big challenge for organizations. I mean, there's these hilarious stories that are already emerging about people that manage to buy a brand-new car for a dollar because someone's gamed Chevrolet's chatbot or a kind of a delivery service that can write you a limerick, but can't tell you where your parcel is.

Mike Walsh (26:28):

So, you hear these stories and these — look, "We tried it, it didn't work. We've disabled the service."

Jim Marous (26:35):

Exactly. The good old it didn't work, not on our fault. Well, it's interesting, Mike, I had a situation recently with Delta Airlines, and because it ends in a good way, I don't have any problem mentioning them, but they have chatbots, they have website listening tools, they have different social media listening tools, and they have humans.

Jim Marous (27:00):

And the integration of those is not as easy as it seems, because in every one of them, they're each trying to perform well by asking the right questions. And they did that very well. In fact, the humans on the airplane asked the right questions, but they each came up with completely different responses or no responses because of what didn't fit in the model that they wanted.

Jim Marous (27:22):

So, I don't think any of the tools they used were bad, except for the fact that they lead you down a path and they go, by the way, we can't give you this answer to actually land. By the way, what we meant by land was at your final destination, but it's not spelled out.

Jim Marous (27:36):

Well, then at the end, a human calls me and the human brings it altogether and says, it appears that you've been using every tool we have available to get an answer. It also appears that we're asking all the right questions. In fact, we're asking it every single time, which is, has to be frustrating. But we're not getting you to an answer.

Jim Marous (27:57):

Well, I've looked over your entire case, I've looked over the discussions you've had, and we've come up with a solution that we think might be good, and they told me what the solution was, and it was a complete refund on an overseas trip. Well beyond what I ever expected or even came close to asking for. Because it wasn't that big of a problem.

Jim Marous (28:14):

But doesn't that kind of look at the challenges of integration of channels, but also the importance of a human aspect to what we're doing in an AI world?

Mike Walsh (28:27):

Yeah, I mean, there's actually so many interesting parts of that story. I mean, the one thing is, I think they actually got to bargain by just giving you a refund because you clearly probably should have got extra for helping them train their model to be more effective. I mean, they were using you basically as a human Guinea pig for their embryonic generative AI platform.

Jim Marous (28:47):

And they may have also in the past where I say, I get in front of a whole lot of people, and if it's a good story, I'm going to mention it. And if it’s was a bad story, I'm going to mention it. And this story that starts off bad gave a great example of what goes wrong, where we have these silos still, we have these old-fashioned silos and try to do things the way we've always done.

Mike Walsh (29:05):

This is in a nutshell, the big challenge regenerative AI today, which is that it's very good at keeping humans deceived of the deeper capabilities of a system, because you feel like at a very surface layer, you're dealing with something highly intelligent.

Mike Walsh (29:26):

But what we are not so good at doing is empowering or plugging that intelligence layer into the ability to plan to segregate to kind of take action even in the digital realm on those inside … from us. And probably for good reason, because we aren't able to audit these things effectively enough to really know that we can trust it with issuing a refund, selling a car, writing an insurance policy.

Mike Walsh (29:58):

We're more comfortable with more deterministic systems with that, because we can actually go and say, if the following rules are met, the following things can happen. But when we can master those highly probabilistic, statistical human-like interfaces, and we can combine it with those more deterministic solve problems, take action, be accountable, that's when the real magic is going to happen, in terms of creating a digital workforce that transforms experiences for customers.

Jim Marous (30:33):

Yeah, it's very interesting because at the end of the day, when you look at where the world seems to be going, the winners are going to be those that can show empathy in a way that says, you look beyond simply the transaction element of what we're talking about, which is, it's hard to do, but as you just mentioned, if you take all the data that I have available in my interactions with this company, and you use it, that empathy seems to come through and trust comes with it.

Jim Marous (31:04):

Again, I mentioned, as my team knows, I mentioned the experience I had in Amsterdam being picked up by Uber and on the way to the hotel, it gives me examples of things I can order to have my meal brought in. And it's all in line with what I've ordered past with OpenTable.

Jim Marous (31:21):

It then says, well, if you don't want this, these are restaurants nearby. And it also uses where I've gone in the past with them to say, here's some events you may want to see, or things you may want to do. It's relatively simplistic, but it takes it out of the transaction and into the engagement. And at the end, you almost get enthusiastic about what the next step could be.

Mike Walsh (31:43):

I've got to be frank with you though, Jim. And I know this is where we're going to go to some extent. But I'm terrified of the idea of equipping these programmatic platforms with too much empathy because I honestly don't think we can handle it as a human species. We are so lonely and disconnected and alienated that the minute that a machine somehow knows a little bit about us and offers us a warm cup of tea, we're literally going to either burst into tears or fall in love.

Jim Marous (32:11):

Good point. Empathy may be the wrong word. Given the human aspect. Yeah.

Mike Walsh (32:16):

The ability for manipulating hairless apes with a little empathy. It just cannot be understated. People are so worried about election interference and deep fakes. They forget just a little bit of kindness connected with a massive amount of data will literally turn us into sobbing blobs of flesh. We're utterly screwed at that point.

Mike Walsh (32:43):

I would much rather cold efficiency from my bots because at least I know where I stand. And then give me a human being when things screw up.

Jim Marous (32:52):

Yeah. So, I'm not sure if I'm leaving on a good note or bad note right now, but let's take a short break here and recognize the sponsor to this podcast.

Jim Marous (33:04):

Welcome back to Banking Transformed. So, I'm joined today by Mike Walsh, futurist, and CEO of Tomorrow. We've been exploring the challenges, opportunities, strategies, and tactics that will define winners and losers in the world of generative AI.

Jim Marous (33:21):

So, Mike, before we were talking about all the things in the AI world and where generative AI moves us. But you've written recently, and you also had a great podcast with Deloitte, talking about the Fifth Industrial Revolution. Can you talk a little bit about what the Fifth Industrial Revolution is and how it differs from the previous four?

Mike Walsh (33:43):

This is an area of some debate and controversy, but I believe we've had four industrial revolutions to date, and they are steam followed by electricity, computation then smartphones and the web. And there's some argument about the beginning dates of those, and whether that was the key technology or not.

Mike Walsh (34:02):

But I think there's no doubt that really we're now in the midst of a Fifth Industrial Revolution, which is that powered by artificial intelligence. And in each of those prior revolutions, there was something being optimized, some kind of efficiency. In the earlier years it was productive efficiency, then it was computational efficiency.

Mike Walsh (34:27):

And arguably now it's really about cognitive efficiency. So, the Fifth Industrial Revolution for me has many attributes. Some people talk about the fact that it's more human centered. Other talks about takes account and other externalities like the environment and other stakeholders.

Mike Walsh (34:44):

But the pivotal part of me, if you take that lens of cognitive efficiency, is how do we most effectively deploy both human and digital resources working together in order to be able to increase our capacity to make smarter decisions at scale?

Mike Walsh (35:02):

And in prior industrial revolutions, like the fourth industrial revolution, which was sort of famously announced at the World Economic Forum, this was the internet of things, digital factories, robotic process automation. A lot of the focus was how do we take our existing capital and really make it work more effectively.

Mike Walsh (35:24):

This worked, I think, but not to the degree that we hoped. I mean, the shocking statistic is since 2004, the kind of productivity rate in the United States has essentially been growing at half the rate that it did for the 30 years following World War II.

Mike Walsh (35:42):

Something is going wrong with productivity for all of these new tools, for all these new technologies, all of these platforms, these investments in cloud and everything else, we are simply not as productive as we should be. And I think the reason for that is we have not been putting enough attention and focus on the question of human beings.

Mike Walsh (36:02):

How do we make ourselves smarter? How do we become more productive and useful? How do we design jobs that are not only better paid, but more interesting, the kind of things we would want to do? And this is fortunately I think the beginnings of a new revolution that's going to change, hopefully all of that.

Jim Marous (36:23):

So, that said, when you talk about the cognitive nature of things and actually the ability to multiply our understanding of things with the help of AI the concept of you can run, take a horse to water, but you can't make a drink hits.

Jim Marous (36:42):

That when you have people very much like the manufacturing area where people knew how to do X, but they really didn't want to or have any desire to move to Y or X plus Y, whatever it may be. How do we actually leverage that?

Jim Marous (36:59):

I mean, in a perfect world, everybody wants to learn more and be able to deploy it in a better way. That's better for me and for the people I work for. But how do we do that? I mean, because not everybody ... people are scared of change. I mean, we see that certainly in the first of the year, every year when people make resolutions.

Mike Walsh (37:20):

I think people are frightened of change. I think unfortunately a lot of it is linked to age. People's willingness to do things differently. Unfortunately, it gets harder, a bit like exercise as you get older.

Mike Walsh (37:31):

And I think, this is a really important lesson for leaders because one of the most important things that we always talk about leadership, about agility and learning actually meets a kind of a natural cognitive resistance as you get older, because it's not that you are incapable of learning, you are terrified of what that means for your current things.

Mike Walsh (37:55):

And this is the probably the biggest thing that we didn't expect to happen, which was that when we talked about AI for the last 10, maybe even the last five years, we really thought the target of the AI disruption of work was going to be blue collar work, people driving trucks, stacking shelves, working on factory floors. And of course, these are jobs that are going to be impacted in some ways.

Mike Walsh (38:20):

But the horrible news that is often difficult for people to accept, it's actually often cheaper to have human beings do that work than get a very expensive industrial robot to do it. It's much more impactful economically to go after the people with higher salaries.

Mike Walsh (38:36):

And these are middle managers, these are I hate to say it, knowledge workers, people who've got high degrees of experience and training, certification, registrations, licenses, people who've got years of experience doing a particular thing.

Mike Walsh (38:54):

It is much more impactful to replace those people because they're already very expensive. And what this means, I think, is that the kind of concept of a knowledge worker has to change, a knowledge worker used to be someone who knew things.

Mike Walsh (39:09):

But in this new environment, a knowledge worker is someone who's able to learn things, who can leverage these new platforms like AI, whatever the latest version of it's going to be, who use it to become kind of a sparring partner in driving kind of their edge of knowledge frontiers and never keeping them static.

Mike Walsh (39:32):

So, the very thing which makes us terrified has to actually be our raison d'etre I think, in the future.

Jim Marous (39:39):

So, that said, what industries, what segments of industries do you see embracing the concepts of the Fifth Industrial Revolution and generative AI and what the potential is the fastest, who are going to be the first movers as an industrial segment?

Mike Walsh (39:59):

Well, arguably, unlike previous industrial revolutions, which were heavily focused on manufacturing, the Fifth Industrial Revolution touches all of us because it is really about that kind of new collaboration between humans and machines.

Mike Walsh (40:14):

So, it doesn't matter whether you're in agriculture or industrial design, complex information industries or in entertainment, life sciences. Anywhere where you've got some kind of complex data or information, whether it's a creative task or discovering a new molecule or the legal professional, the finance profession, they're all going to be changed in different ways, but at the same level of impact.

Mike Walsh (40:43):

So, unfortunately, it's still early days, I think for almost all of these industries, but I think particularly the ones where there's a high degree of knowledge required, these are going to be the first in the firing line. And this is already starting, there's already been massive changes in the employment base, in financial services. This is already starting.

Mike Walsh (41:03):

A lot of it was done under the cover of a potential recession. But be under no delusions about it, AI is driving this. It's already happened in the technology industry, all the massive hires that were made, especially around sort of mid-level managers during COVID where basically Zoom meant we had sort of a proliferation of people whose job it was to check on other people were doing their work or not. All those people have now been laid off and they're gone.

Mike Walsh (41:31):

So, in any sort of area where there's a high degree of complexity of knowledge and data already, I think even though these tools that haven't hit primetime yet, leaders are asking themselves, are we at the right scale for an AI powered era?

Jim Marous (41:47):

So, that said, I think most people in the world today are worried about being replaced by the machine, the AI tool, whatever it may be. How do workers prepare themselves for the future to make sure they're part of it as opposed to being on the outside looking in?

Mike Walsh (42:07):

Well, I think, you have to sort of put into context the speed by which this is going to happen. There's no imminent job apocalypse. These things always take — you know how they say that they happen slower than you think, and then they happen faster than you expect.

Mike Walsh (42:26):

So, there's kind of a ramp up phase to this. So, there's no imminent decline. And in fact, those firms, I think that act too quickly to get rid of people are going to be hoisted by their own batard. I mean, my favorite example of this was Sports Illustrated. I don't know if you are following this, there was that whole scandal about all of these AI generated writers, and they even had fake profile pictures.

Mike Walsh (42:54):

And this started because the CEO of that group had made some very public comments about, he thought that most of his people were useless and weren't doing anything. And obviously he kind of pushed aggressively to have them all replaced by robots.

Mike Walsh (43:11):

The irony after all of this is that most of the writers kept their jobs and he was fired. Which is, a perfect example for me is that there will be casualties of the AI revolution. They might just not be the people you expect, and they probably a lot more senior than you'd think.

Mike Walsh (43:26):

So, I think one of the most important things that you have to bear in mind is that context that this is coming, but it's not coming necessarily as fast as you would think. Humans are still very complex, and we do very complex work that are difficult to get machines to do.

Mike Walsh (43:44):

But if you are not today already experimenting with using these tools, becoming comfortable with them, recognizing at a very deep level that your job is not to work. Your job is to design work, to find a better way of doing the work, of finding tools to change the nature of the work.

Mike Walsh (44:04):

This was as you mentioned before, when we last spoke, back in 2019, this was the big theme of my book, The Algorithmic Leader. I mean, all these things hadn't happened, but back then, even though I wrote, "If you could imagine a world which AI becomes prevalent, the most useful thing you can do is not the job as it was originally designed to be do, your job is now to think about how AI changes the nature of that job." So, that advice that I gave back then, I think still, if anything else is even more true now.

Jim Marous (44:30):

Yep. It's interesting because there's so many dynamics here and as you said, it gets down to how fast and how cognitive can you be as a human, but also the technology behind that to move forward. What excites you the most about the near-term future?

Mike Walsh (44:52):

I think the most exciting thing for me is that we are really now equipped to be able to learn anything we want in a very precise and accessible way.

Mike Walsh (45:12):

And the reason for that is if you think … I always used to imagine that if you had the ultimate luxury, you'd have a private tutor. In the sense of the wealthy aristocrats of Europe, you never went to school or college. You basically had your own personal tutor who you'd do a grand tour of Europe, and they'd travel with you and you'd be somewhere in Rome and there would be someone lecturing you on classical history or Latin or Greek as you were standing in the Parthenon.

Mike Walsh (45:41):

Or if you were a great leader like Alexander, you literally had Plato as your private teacher, who was schooling on the classics directly. Anyone can have this now, you can basically use in an early form, these AI tools to kind of be your guide to kind of a world of infinite knowledge.

Mike Walsh (46:02):

If you have the curiosity and you have the questions and you have the passion, the ambition, and the drive, there is no limit to what you can learn. Because rather than just sort of mechanically reading through Wikipedia, you can now ask questions and follow up questions and use that to basically navigate kind of very complex trees of knowledge.

Mike Walsh (46:26):

And I think if that's where our starting point is after 12 months, what's this going to look like in the next 10 years? And of course, this is I think a plea that we need to transform education now, but it's also a very positive thing, which is even today, a lot of the barriers towards self-learning are going away.

Mike Walsh (46:48):

So yes, it is frightening learning new skills, but if you even have a tiny modicum of curiosity and a willingness to kind of embrace new ways of doing things, there's never been a better time to do it.

Jim Marous (46:59):

It's interesting learning how to learn. I mean, it's very interesting that way.

Mike Walsh (47:03):

This is what we're talking about metacognition, thinking about thinking.

Jim Marous (47:06):

What are you most fearful of given the, I say short-term with regard to AI and generative AI-

Mike Walsh (47:13):

There's two things that terrify me, one we already touched on, which is the kind of misuse of a mode of technology. Because I say humans are psychologically very weak. There's been a lot of research done on this, we can be nudged, we can be hoodwinked, we can be conned. And we also tend to anthropomorphize, and we tend to worship things as gods.

Mike Walsh (47:38):

So, I think tech geeks are particularly susceptible to this, creating something that seems like a digital god and then starting to worship it. And I don't mean this in a religious sense, but really just kind of-

Mike Walsh (47:55):

Or political.

Mike Walsh (47:56):

It could be, it could be political. But I'm sort of specifically thinking, that it won't be very difficult and it's going to happen very soon to create something that we consider a GI, whatever that is, and then assume that it has feelings, and that it needs to be given agency and autonomy or citizenship.

Mike Walsh (48:13):

I mean, it's ridiculous. And it's not that this thing won't be able to trick most of the people, most of the time, the question you have to ask yourself is this obscene? And I'm not trying to defend human values, but it's just that to what purpose is it to create something that is able to simulate consciousness or to trick people into thinking it's consciousness conscious.

Mike Walsh (48:38):

That doesn't make us gods. It actually just means you've now created something that's less useful than a tool. It's like a hammer that wants to talk to you rather than hit nails. So, that's the first thing that frightens me.

Mike Walsh (48:48):

The second thing is that out of fear and ignorance, we overregulate something before it gets started. And in doing so, we plan to the hands of the large companies and platforms that have got an early start on this, that have an interest and create a regulatory barrier to other new players entering or leveraging open-source technologies. Those are my kind of twin fears.

Jim Marous (49:10):

Very interesting. The second one is interesting because on one hand, regulation is needed, but to what degree, and again, both of them play off the fears of people. The worshiping a non-human thing.

Mike Walsh (49:28):

False idols.

Jim Marous (49:29):

Yeah, exactly. And then the other one is regulating that way, but it's the fear of the unknown.

Mike Walsh (49:35):

Fascism and religion are kind of the twin traps of human civilization. And if you kind of look at the span of history, any new technology often falls into one of those traps.

Jim Marous (49:50):

So finally, beyond simply get started, what advice do you give businesses right now to get prepared? Because I think the biggest challenge is there's a lot of priorities out there, and you can test yourself into a hole and never move forward. It makes you feel good that you're doing things, but you're not really.

Mike Walsh (50:11):

No. And I think, Jim, let's be frank here, the time for screwing around and teaching executives how to write prompts, having a workshop on how to use generative AI, the time for that's passed. I mean, frankly, if you haven't done that as an organization now, tell people to go do it at home. Don't waste time on running workshops on using ChatGPT. It's just moronic.

Mike Walsh (50:40):

Really this is the point where people need to be getting together and going, "Okay, not just generative AI, what is our strategy for putting AI at the core of our business? And if we don't do it, what is the consequences of one of our competitors putting AI at the core of their business? What would that look like? How would that impact us? Are we okay with having the valuation profile of a company that isn't AI leveraged?"

Mike Walsh (51:08):

And honestly, for some companies, that may be okay. It may be that we are prepared to take this valuation profile because we are not willing to risk the capital required to chase the kind of a hundred X version of it. And that is an entirely reasonable strategy for the vast majority of firms, as long as it's a conscious one.

Jim Marous (51:30):

Yeah. Mike, it's always very intriguing, always fun. I'm lucky this time because I actually know the next time, I'll be seeing you, you're going to be at the Financial Brand forum in May in Vegas and going to be doing a session there. And I'll look forward to sitting down with you and say, "Okay, even though length of time between now and then, what's changed in the world."

Jim Marous (51:52):

But it's exciting. And I think, the opportunity is so great. And the not losing the desire to learn is one of the things that keeps me going because I don't want to play my age, didn't want to do it when I was younger. I certainly don't want to do it when I'm older.

[Music Playing]

Jim Marous (52:13):

But it's interesting because there is so much out there and now you have tools that, as you said, can amp up that learning very quickly because it's all at your fingertips. It's kind of exciting. Mike, thank you so much for being on the show today.

Mike Walsh (52:28):

My pleasure, Jim.

Jim Marous (52:30):

Thanks for listening to Banking Transformed the top podcast in retail banking and the winner of three international awards for podcast excellence. We appreciate the support we have received to make this endeavor a success. If you enjoy what we're doing, please take some time to show some love in the form of review.

Jim Marous (52:47):

Finally, be sure to catch my recent articles on The Financial Brand and the research we're doing for the Digital Banking Report.

Jim Marous (52:54):

This has been a production of Evergreen Podcasts. A special thank you to our senior producer, Leah Haslage, and our audio engineer and video producer Will Pritts.

Jim Marous (53:03):

If you've not already done so, remember to subscribe to Banking Transformed on both your favorite podcast app and on YouTube for more thought-provoking discussions on the intersection of finance, technology, and leadership.