The Human Code
The Human Code" podcast unravels the intricate blend of technology, leadership, and personal growth, featuring insights from visionary leaders and innovators shaping the future. Host Don Finley dives deep into the human stories behind technological advancements, inspiring listeners at the crossroads of humanity and tech.
The Human Code
Building Trust in AI: Holistic Approach with Michael Proksch
The Human Code: Bridging Technology and Humanity with Michael Prokopis
In this episode of The Human Code, host Don Finley takes us on a journey through the intersection of technology, leadership, and personal growth with guest Michael Prokopis. They discuss autonomous driving, the evolution of AI, process mapping in supply chains, and developing future leaders adept at navigating both technology and human-centric roles. Michael delves into his background in technology and strategy consulting, and how he applied these insights to innovate in healthcare supply chain management at MD Anderson. This conversation is packed with insights on how technology not only solves today's challenges but also shapes the leaders of tomorrow.
00:00 Introduction to The Human Code
00:49 Guest Introduction: Michael's Journey in Tech
01:03 The Impact of Technology in Daily Life
02:20 Autonomous Driving: The Future of Transportation
04:14 AI's Evolution and Challenges
06:39 Data Privacy and AI
09:05 Future Leaders and Technology
11:53 Supply Chain Leadership and Development
16:01 Process Mapping and AI in Business
22:38 Balancing Human and AI Collaboration
27:25 Career Advice for the Future
30:33 Conclusion and Contact Information
Welcome to The Human Code, the podcast where technology meets humanity, and the future is shaped by the leaders and innovators of today. I'm your host, Don Finley, inviting you on a journey through the fascinating world of tech, leadership, and personal growth. Here, we delve into the stories of visionary minds, Who are not only driving technological advancement, but also embodying the personal journeys and insights that inspire us all. Each episode, we explore the intersections where human ingenuity meets the cutting edge of technology, unpacking the experiences, challenges, and triumphs that define our era. So, whether you are a tech enthusiast, an inspiring entrepreneur, or simply curious about the human narratives behind the digital revolution, you're in the right place. Welcome to The Human Code. In this episode, we have the privilege of speaking with Dr. Michael prox. And accomplished expert and leader in creating business value with data and AI. Along with being the celebrated author. Of the secrets of AI value creation. Michael has an extensive experience working with fortune 500 companies across Europe, Asia, and the us offering innovative solutions and driving transformational journeys. His unique blend of expertise in business analytics, technology and psychology has earned him recognition as a thought leader and sought after speaker at conferences worldwide. Today, Michael and I will share the secrets behind holistic AI value creation and why integrating technology analytics, business data and psychology is essential. How to build trust in AI systems to ensure successful adoption and meaningful value. His vision for the future of AI and how to maintain a human centric approach. Can lead to greater innovation and job satisfaction. Join us as we dive deep into these insights and more with Michael.This episode is packed with valuable takeaways for anyone interested in the future of technology and leadership. You won't want to miss it. we're here with Michael Proksch. I gotta say, Michael, it's an absolute joy to have you on the show. one of the things that we've connected about is a passion for AI together, but at the same time, you've got a, an aspect of it that you've dived deep into AI value creation and like what it actually takes. And I know from like the customers that we talk to, the people that reach out to us, that is the biggest question, right? They want to implement AI, but at the same time, A, where are they going to do it? And how do they do it responsibly? So, once again, thank you for being on the show. my first question to you is what got you interested in like the intersection between humanity and technology?
Michael Proksch:of all, thank you very much for the invite and I appreciate it. I really like your topic and I really appreciate, you finding me. So that was really cool. and thank you for actually buying my book before we get into it. and actually reading it, I was surprised that, you came prepared and you really You know, really want to talk about the topic. so yeah, the topic of AI value creation has been one of mine for a long time. Actually. It's not everybody now focuses on it, but has been mine for, a decade. actually it's about, this, the struggles of many people, of how do you create value with the data analytics, and I did too. So we all had the same struggles and I. The only thing I did a while ago was, to not, trying to find my mistake in, where do I have my own mental failure? And that was when I was focusing on analytics and technology. that was, back in the days, when there was nothing, we have to build everything ourselves. And I was like, okay, I don't create the value I actually want to create. So what am I doing wrong? And then we started mapping out the other factors for value creation. And we figured out it's yeah, analytics and technology, and you have the business part and you have the data part and you have the psychology or the humanity part. and then we realized, wait, we do something wrong. Most companies do it wrong because they're focusing on two out of five and they think, okay, good. Somebody else is going to take the other three. and that's usually, what's not happening. So we miss the other three and that's where it starts to fail. Or we focus on the other three or two out of five, or, many companies hire people in all of those five sectors, but they never talk to each other, they have their own agendas and they have their own optimization function for their own different areas, but if they don't interact, if there's no, intersection between those five, there's no value created. It's really about, the multiplication of those five factors that eventually create value. And humanity is an important piece of it. in the book we actually talk about the big three factors like business opportunity, feasibility, and adoption, right? And adoption has everything to do with humanity. If we don't build it for people and show people the value they have from it, Why would they use it? And that's something we always seem to forget. We go and say, Hey, look, there's a great tool. There's something amazing for you, but we don't, pick them up from where they're coming from and don't help them to see their own value by using it. Why should they use it? How many dashboards, and I know you, you have the same background here in software and development and AI. How many dashboards have you personally built that actually nobody used? I'm
Don Finley:If anybody says anything, if anybody says zero, they're lying to you.
Michael Proksch:that's
Don Finley:Because we've all created something that, we think is amazing and then all of a sudden it gets in front of somebody and, we missed it. Or, we just created it and it was You know, it's used for a day, it's used for a week, and you go on. But I think you, you hit on three of the five value creations. remind me of them again.
Michael Proksch:So in general, it's the five factors, technology, analytics, business, data and psychology.
Don Finley:Ah, okay.
Michael Proksch:And then, and, in the book we read it slightly different. but it's, eventually it's about, how do you combine all of them? How do you make them all, Because all of them create their own hurdles and challenges. And that's actually part two of the book. those are then the five that come up there in different ways. Where we talk about, okay, how do you focus on And it has its own challenges. And how do you focus on, actionable data? Which creates business opportunity. And then how do you focus on, trust, which is psychology. So if I come to you and come to you with a wonderful product, but you don't know me, so I could use that product and AI for you or against you. and that's the biggest problem with dashboards. Many people actually struggle with the, not just trusting the, the technology and the data, that's one thing. And, not being accurate on, helping you to actually do your job. But what do you actually want to do with it? Like you measure the performance of people. Or you use it to help people to have a better performance. same data, same dashboard, totally different intention.
Don Finley:Oh, exactly.
Michael Proksch:And that's where we're writing in the book about like, how can you start from a different perspective, right? I always try to create trust first and show people that I'm really benevolent. What I say I'm doing, I'm actually going to do. And then, we're doing it, but I need them to trust me first. Otherwise there's no value I can actually create. And if I, I don't know if I misuse that trust, then nobody ever wants to work with me in that organization again. So it's it's all about trust because, AI is new. People are scared. It's not necessarily, everybody's embracing it because people think, okay, it can take over jobs or at least pieces of it and tasks, Which might actually help. it can create, or it can take over tasks people don't like to do, which is great. but is it going to be used against you or for you, right? Does it help you to. make your commission or is it going to help you help the organization to reduce your commission? Totally different intentions, same tool, right? And I think that's something many organizations haven't really figured out. Um, how do you get the people on board and getting you, to trust you and, the, help not just the organization grow. We talk about three different stakeholders that we usually try to serve, right? We try to serve the organization, we try to serve the customer, right? But we always forget the people. So we also get the people that are working with it, the stakeholders that are supposed to use it, if we miss that piece, then no adoption, that's in a very emotional factor, And we usually forget as technologists and analysts and, to include those pieces. And that's the human part.
Don Finley:And I, I remember the first time that really hit me like flat in the face. I went from the, in banking or the financial services market to the vending market, so like vending machines and who fills those machines. And in banking, you had regulations. And so everybody did the work that they needed to do because there was a penalty if you really didn't do it, but in. In vending, there, there was no regulations, there was no space. And so you had to know what that person's intrinsic value was from utilizing your way over the way that they've previously done it. how do you see the value creation working for the individuals, for that community that's part of it, for the users, and helping them over that cliff of building that trust?
Michael Proksch:I think the first one is that we need to involve them in the process, to build it because eventually it's their tool, whatever we build, it's for, it's has to be for them. Otherwise it's not gonna create any value. When people get involved and see that they can influence the outcome and how it gets used, you create trust, right? Even before the tool starts to go in production, I don't want to call it, get used. and I think that's something I've been always trying to do. After I learned my failures from my own failures to really include the people on that journey, right? Because there's, there's one thing you say and the other thing you do. and if both goes together, you create trust, right? Yeah. Integrity, That's what I at least try to do. How do you do that? I work with my customers very closely, and you as a CEO of a software company, so you have the same struggle,
Don Finley:I look, I really appreciate you asking the question. Um, I definitely agree with you that we need to be looking out both for the community that is using this, the organization, as well as our customers. and if you can align the goals of all three of those together, that creates a winning product. I think, I think I learned early that I can't ignore the users. and this is from an enterprise like business B2B. type of situation, right? Because you, when you're delivering enterprise software, you're seeing it that your enterprise user is different from your enterprise buyer. Yet the benefit of your system, the benefit of what gets implemented, the actual value that you are creating, everybody is necessary. And so from a stakeholder standpoint, we look to do the things like you're talking of bringing in users early, bringing in people to get feedback on what is being created, getting them to be the ones who are helping to determine what is actually necessary for it. and at the same time. The way that we do AI implementation now is that we work at two fronts. We work at the board level or the management team level. And then we also work at the individual level. And so at the board level, you're planning out the larger implementations, but then on the individual level, we're creating workshops that allow people to say Hey, here's how we're utilizing. AI in our function because they, people are using it is basically what we found out. They may not be telling you that they're using it. And that's a problem that every organization face from a risk, perspective. But when you can pull it out into the light, you can set aside security policies that make sense. You can teach people the idea of having, whether it's in a local LLM versus a cloud LLM, and then also understanding Hey, am I giving away my data for training? Or am I retaining some privacy to that as well? but I gotta tell you, like most of the time, some of the best ideas for how to use AI come from individual contributors and how they have tried to, improve upon their work product as well. and so you gotta change the conversation is really what we've been working on. And allowing, yeah,
Michael Proksch:And I think you're totally right with that. So I think, humanity what they want to use the technology for is the way to go, and not,
Don Finley:with all of the work that you've been doing over the last 10 years, right? And if we really focus on the value creation side of this, plus the humanity component, you might have an, I want to hear your take on where this goes in 10 years.
Michael Proksch:I think, something, Elon Musk is doing very well. And I think that this, showing his success, it's the involvement of humanity in building the product. So it's yes, he has great vision and he's going, he's, Doing things that everybody feels too or, much out of the comfort zone or, say, he, he took great risks in the past, but, building, building a self driving car over the last, say, 10 years with the help of people. So he didn't claim he has it. So he had the help of all the customers and building it. So he did it that way. or the, you know, when we're talking about, Twitter and now X, right? He's doing that too. So he's listening to people getting their feedback and he's. pretty transparent in what he's doing to a degree. It's not, always, helping, but it's he, he does it exactly like that. And he's been pretty successful with it, now with the new robotics and, he's sharing the, where he's going and what he's doing. And it's pretty impressive for him. not just him, the others too, but it's he's really focusing on, on value creation in all different kinds of ways. And and I think that's a good example because a lot of people invest in Tesla and, looking at the Tesla stock did over the last. 10 years. It's incredible that, he showed benevolence. He showed integrity. He's following up on what he's doing and saying. so I think he's showing in the spotlight that this approach is actually working and including people in the, in that process. He's answering people on Twitter, that complain about his product, and I'm like, yep.
Don Finley:And I do remember there was like, Certain aspects of people would say this would be a great feature in the car. And then a couple of weeks later, like it's released type of approach. And I can tell you from being a product manager, that would be absolutely a disaster to have a CEO in your product pipeline that, that deep. But I do agree with you that talking to your customers, getting ideas from them, understanding how they're using the product is how every CEO needs to be behaving. there's.
Michael Proksch:Yeah, I don't usually judge people. So it's for me, I can highlight the people that are very successful with a specific approach. And that's just one, right? Steve Jobs was, I don't want to say the same because he did it differently, but he Communication as well. So there's a very way, like a specific pattern of, how successful people are, successful. And I, especially with the book, I realized that when we were writing the book and we already had the concept and the ideas and. We thought, hey, we were, we were pretty unique and thought we, I thought many years ago, I cracked the, the success formula. And then we started talking to all the contributors of the book, You have, so many, CDOs from big companies as, contributing to the book. And we realized they have been doing it for many years. That's why they're successful. Yeah. Okay. Okay. Good. yeah. Okay. And they shared their learnings and their, their stories and how they did it and what they, and what didn't work. and it was, we seem to be very aligned on that. I know, I
Don Finley:that you got from these CEOs or like how you were coming about creating the value creation pipeline?
Michael Proksch:mean, there, there's no specific story I can actually highlight. I was more, surprised in a way that they all shared the same story. Like that, that, that holistic picture of, and we talked before show, and we talked about the book and you've been reading it and you said, where you didn't seem to be surprised. It's oh, there was not much, whatever we shared, you seem to be like very on. on my page, and that's something I experienced with successful people in that field, that they all agreed to this stuff, but there was no book. there was no course, no class in college you could, go to. I haven't seen it. I was like, okay, wait, why is there no book about it? Why is there no, why is there no, nobody who's like pointing that out? And it's that's why we wrote it, right? It's okay, that, that was missing. and then, um,
Don Finley:there. It's in front of us. and at the same time,
Michael Proksch:and ma many
Don Finley:outlined it.
Michael Proksch:That's exactly right. That many people that we talked to, right? they apply things that are like 40 years old, right? David Porter, Michael Porter and other, famous people that, that we know the approaches for many years. and we haven't, we seem to drop all the knowledge we have from the past. As soon as we hear ai, everything is new. So let's forget everything we know. It's no, way, why would you, why are you doing that? Transformation, change management, all those things are even more important for that, technology than it has ever been. And we seem to forget that, right? We seem to forget all those things that we have learned or they require just nuances on adjustments and a slightly different perspective on it. But in total it was. Talking to those people that they had all those approaches used for many years. And then they brought them together to do those intersection between all of them. And that's why they were successful. And for some reason, that wasn't a big surprise. They just, they always said, yeah, that's great. There's no book. Okay, good. And you said the same thing, right? So you bought the book. and I was, surprised and you said you put it down, and haven't read in it for two weeks because I'm sure because, after the first hundred, 50, 200 pages, you said, okay, I know all, most of that stuff and I'm doing it, you
Don Finley:Let's make this straight. I didn't put it down because I was like, I know everything. I put it down because I had to finish up another piece that I was studying.
Michael Proksch:No, which is fine, it's all good.
Don Finley:yeah, I agree with you, what you're talking about, I'm what I was surprised by is a how simple it was, but then also what I came back to realize is I've been having those same conversations with people where like when I'm, we end up doing work with other software development agencies as well, right? Like they come and they hire us to be their AI team. Or at least to help them with that AI implementation and figuring out like what projects to go after. I'll be on the phone with the like vice presidents of data analytics for these companies and I'll be like, guys, this is just another digital transformation. It's what it is It's the same stuff that you've been doing. it's a new technology. It's nothing to be frightened by. You still have to justify it in the same ways that you've been justifying in the past. You still have to evaluate it in similar ways, but it's a different type of tool, right? Like it, it's a shovel that knows how to make a decision.
Michael Proksch:Yeah, I think it's just the new, the new application, the new capabilities and risk that people, start to, start to realize, and which is good. the book eventually is just a guide of, bringing all the bits and pieces together and not forgetting one. Because it's a complex, it's a complex field, right? Especially if you're coming with an education in, computer science or, so you're usually new to the other areas. So, and then you usually, and I don't want to judge anybody, but it happened to me. so where you start to, to put a more importance on things, because, you don't put the importance of things that you don't know. not. So you start to, to reduce your efforts in those areas and that's where everybody's. is failing, right? But actually you need to pull them up at the same time. So it's the journey you go through, the initiatives you build, it doesn't help you if you have, the best technology, invest 5 million and don't have anything else built. So the value you create is zero. So it helps the most if you take your 5 million and allocate it equally over, the, all those five factors you pull up step by step, or you don't even need to start at five, but it's you need to. allocated at the same time and then, go step by step and to create value, right? Slowly, it doesn't happen overnight. Now, the most of the journeys we talked about in the book, right? to get to where they are as AI achievers, it took them up to 20 years and more. So they started by, collecting right data and they did that 20, 25 years ago. John Year started collecting data with the GPS data many years ago, and now they have self driving tractors based on the GPS data. They collect, but it's like the experience they made over time. And, they made all the mistakes and they overcame all the challenges and learn from it. And, where, you know, You that, that's another reason why it took them so long to actually, overcoming all those hurdles and learn because there was nobody like 20 years ago when AI was still, I would say, very new.
Don Finley:look, like we're talking neural nets and like the stuff that we saw in 2013 around like image classification, I think was like the big thing that kind of jumped it because we figured out how to use GPUs. But like that math behind all that. Came from like the 70s, didn't it? It was, I remember studying it back in college when I was an undergrad and like doing a little bit of what is it? the numeral, numbers. Recognizing numbers, right? And that was like one of the use cases where you're like, Oh my God, neural nets are going to be amazing. But that took hours to process on I 86 platforms back in the day. And then in 2013, 14, we saw, a jump there, but to have that foresight of Hey, we need data. We're going to need compute. Do you have any recommendations for people that are looking to implement and what roadblocks that they should be wary of in their quest of that value creation and an outside of making sure that you are adopting all five of the pillars?
Michael Proksch:I think, we wrote the book like that. so the first part of the book is, the AI value potential, understanding how it can actually create value, right? And we show a couple of AI journeys, and that they took 20 years and how they, how they walked that journey. And, that it was a marathon and not a split. Right. and then we talk about, The second part is where we talk about the factors of AI value creation, right? So where we said it's business opportunity, it's feasibility and adoption. Those are the big three factors we try to, we try to optimize towards to create value. many organizations and Um, focus on that's just one of the mistakes. And we talk about a lot of mistakes in the book, because those are the things, if you try to avoid, if you avoid the mistakes, you eventually create value. So if you create the risks and the costs and the mistake, the mistakes with that, the costs of failure, you, you create value. So it's going to take a while, we talk about all those, mistakes and things that we have seen or done or, ourselves. That one of the things was, when we talk about use cases and I have those questions all the time, or what are the use cases and, there are those wonderful use case catalogs and mostly by industry, right. And that's and I always have to, uh, a smile on my face because, one of the things we learned from Michael Porter that, in order to be successful, to create value in your industry, you try to create a competitive edge, right? So if everybody's doing the same use cases. How do you create a competitive edge with it? The only thing could be that you're faster and the pioneer and you did it five years ago. So then you created, then it became an industry use case, but you already have it. So in order to create value or business opportunity, you need to create something unique. something that supports your organization, something that supports the value creation of your organization, something that is the unique value proposition of your organization. That could be in pricing, that could be in, you wherever you want to focus on, it could be the fastest delivery that, whatever it is, that's where I can actually help you. And that usually you don't find in the use case catalog. So, and then, or you can, you can buy it, you go to Salesforce and wherever, they're mostly readily available already. because that's why it's an industry use case. The next one is when we talk about the ways AI can create value and, they, they have an increasing risk, right? Everybody wants to create products and services, which is great, but that's the highly, that the highest risk. You can actually, take and creating something because that you need to, there are all the risks of business opportunity, feasibility, transformation, adoption, and there it becomes a product market fit and building, building the right marketing, building the right, it is so complex that's why we have such a high failure rate in the startups, right? Because that's The highest complexity you can actually take on an, on AI projects, right? So where you can start is in business process optimization, low hanging fruit, right? It's already there. The processes are already mapped out. The actions are there. You just adjust the actions a little bit, that they are a little bit more performant, not the highest opportunity, right? But definitely the lowest.
Don Finley:fruit.
Michael Proksch:The lowest feasibility risk, the lowest adoption risk, and people ask me where would you start when you would start again in an organization? And I always say in the mailing department because that's how you become a millionaire, you know You hear that all the time Either the dishwasher or in the mailing department where people usually start their career and same thing for me I started many times I started in the mailing department optimized mailing Because that's usually a cost factor nobody cares about because it's a cost that always comes. it's a cost factor that is accepted. It's, but it's usually significantly high enough that if you start saving there, you can show some low hanging fruit. And it's a process that's already mapped out and people are involved. There's no adoption risk. Usually, you feasibility, you have enough data about your customers that you can, say who to send stuff to or not. that's my. my go to case when I talk about, process optimization. And there's meanwhile companies that can help you with that.
Don Finley:Which is amazing. Like we've been teaming up with, business process automation companies, and they do the analysis on your process.
Michael Proksch:yes.
Don Finley:And then they're like, Hey, and then we'll review the results of that and be like, all right, so this is a case where you could actually take agentic AI and move it. And this is another case where you can put like an AI platform in here as well, or just your standard expert system. Kind of plays in that role. And those are nice, easy wins for organizations to go after.
Michael Proksch:And you're totally right. So I was really lucky to get the CEO of KYP, Adam Boyack, Dr. Adam Boyack, to contribute to the book on exactly that chapter on process optimization. So he was, head of, automation process optimization at Capgemini globally. So he was very kind to, to contribute to it. So we have like contributors that actually, that really focus on those chapters and that are their chapters because that's their background and that's their focus. but like going from there, those are the two extremes. And then in the middle, you have decision augmentation where you create information beyond what people already know. And, that's requires some transformation change management has to feasibility risks and include it and then, automation where you start to automate actions and that usually creates more risk, in different ways. so we talk about those four types of, value creation and, where to start and where to go and what are the risks and different areas. And then, the second part of the book is really about, how do we come, overcome the hurdles, right? How do we come, the hurdles of a specific project, right? And that's what are, how do you evaluate the value of data? Right. And then we started and I actually, I started from a concept from, professor Akoff, famous system, professor from Wharton, school, and he wrote it in the nineties. 80s actually, in 1989. From Data to Wisdom. Wonderful article. So I read it at least 25 times and I still find new cool things that I'm always surprised about. Oh, that's cool. So
Don Finley:So that's what, is that where we get the whole data to wisdom pipeline?
Michael Proksch:Yes, and that's that's, he came up with that. And I use that in presentations and conferences all the time
Don Finley:I use it all the time too. And I never knew where it came from.
Michael Proksch:and it really explains if you, in the book we have a whole chapter from, how do you go from wisdom and knowledge down to data to actually evaluate the value of it. and it really, going up and down and left and right, top to bottom and so on. So it really explains so many things. You also, you that, and that's very interesting when people always ask me, okay, what is the difference between machines and AI and, humans and AI? And I use actually his approach. in the book, he is actually talking about, knowledge and what knowledge means and what wisdom means and, how you understand AI. In order to get to wisdom, you need to understand knowledge and you need to create something that is related to growth and development in order to be wise. And that's why he says, okay, machines are never going to be able to create wisdom, but, are able to create knowledge. And that's exactly where we make it, where we make the mistake when we talk about, artificial and human intelligence. Because usually wise people are knowledgeable. So we misunderstand the important, the definition of intelligence and think about, and wisdom has to do with intelligence, but that's not true. Wisdom, intelligence is related to knowledge. So artificial intelligence can gather lots of knowledge by putting something new together, something that helps humanity to grow and thrive. I'm on the page right now, at least with ACOF that. This cannot be done by machines because when you think about an email you receive recently, a lot of them are written by chat GPT or some generative model, but people use to write an email. and when I see the patterns in it, like best regards is usually not capitalized. So that is usually a good hint that, it was written by chat GPT. and then, but immediately I think, okay, that person must have been lazy. using Chatchuby, you can't even write your own email. So I was like, I immediately devalue, although the grammar is fine, everything's great, but I have a weird feeling about, okay, that's done with Chatchuby. So now, in five years ago, when somebody made a mistake in an email, I was like, okay, that person didn't even read it again and sent me an email. And today it's oh, that person made a mistake. Oh, I put so much effort in writing it himself. It's you know, I was like, Oh, okay, cool. So that email is, human. And that's something I realized myself. that's, that was, exaggerated and, but it's we
Don Finley:ever seen, continue? Sorry.
Michael Proksch:yeah, no, it's all good. When we started to talk about, value creation, it first created value in writing the perfect email, but do we value that email now the same way as it was? before, that we know it was written by machine. So I don't know. I also like to see, recently I saw one of the videos on LinkedIn when, then they're like those RoboDocs, right? And they're pretty impressive, What they can do and how they can jump and do things. and I saw one of those videos where one of those, you know, people that work for the company, kicked the RoboDoc. and I was like, okay, so they are so used to that machine and he didn't push it and he didn't want to show that, it could balance kicked it. And so it's okay, is that where we're going to go that when we have it all around everywhere, we wouldn't kick another human to show that this person can balance. But we kicked the robot off. So I, is that something we, is that human? is that something we,
Don Finley:What you look at it, we do it on both sides, right? Like we will both kick the robot, but we'll also personify the robot, right? Like we'll give it characteristics. And I know I did this when I. I was breaking. I can't remember the year, but it was probably 2014, 2015. And I was getting back. it was early on in TensorFlow, as far as that goes. And I created a, a. An RNN that basically read my text messages to my girlfriend at the time and then created a persona that was her, right? And so it would input my text message and it would output her response. That was the basic thing. But the structure of it was the only inputs that the AI knew were the alphabet. So it just knew characters and then it had to build up that knowledge and it was terrible. But one time it responded with, I was like, what should we have for dinner? And her response, the AI's response was, I miss Vincent. And Vincent is the name of my dog. And it hit in a way that it was like perfect for her personality and what she would have responded to one of my questions with. And I just felt myself putting that human characteristics into the machine. And so I think we, we do, we kick the robot. I've seen the Boston Dynamics where they are brutal on showcasing how well these things balance. and I hope that we're not going around kicking robots. When the day comes, but we also show capacity to do the opposite and along the lines of going to that data, the knowledge and wisdom aspect of it. I feel we may fall prey to prematurely identifying something as wisdom from a machine when it is just a really nice mimicry.
Michael Proksch:that's right. And I think, and then it's knowledge, right? Or going to, going to chat GPT or, ask for, show me a green cat. Knows the concept of green, knows the concept of cat and can put it together, which is fine because it has, created a blue cat and white cat, black cats. can create that, but ask for the future of humanity. So what, how does that look like? like where, because it hasn't been trained or it just gives you back what, we already know from Star Trek and Star Wars. But it's nothing that is, in some way,
Don Finley:Now, LLMs are basically like predictive machines, right? that's effectively the structure that goes into it. but we get a
Michael Proksch:far as I'm not the expert on LLMs, but. But as far as I understand, yes, if you listen to Elon Musk and he talks about it, that's the, it's going to take over the world, he might know more than I do. So that's the other thing, when people always ask me, okay, what's going to happen in the next five years? I said, okay, if you would have asked me five years ago, I would have been totally wrong with whatever I could have told you.
Don Finley:I was wrong two years ago, and I'll be the first one to say that. I thought we were gonna get self driving cars that we were also going to be down the path of like automation around like blue collar work was really where AI was going to play in. But the last two years have just shown that everything's fair game today.
Michael Proksch:Yeah, so that's the, that's part of the second part of the book. And then talking about like how, what would, how can it become actionable? What is actionability? How can they ask, how can AI help with that? and then we talk about, trust. How do you create trust? We talk about, the nuances of AI. So how is AI different than software development or software in general, or dashboards? And then, we talk about, managing AI's decision making, right? we usually overestimate what the AI can do, but it usually really just optimizes one outcome, right? If you go and optimize, your YouTube video towards, conversion, it doesn't mean that the person actually buys it. they just clicking on it and come on your website. But, or how long they actually watch it. It could be, a 2-year-old that's sitting there and doesn't know how to push the button. So I was like. there are so many misconceptions we make about AI and, we're focusing on, bias, generalizability, then if you overcome all those little challenges, that really can crash your project, so I call them little, but they're actually pretty big because they are the unknown unknowns usually for a lot of people. So how do you have then one use case, how do you integrate it in an enterprise that is like 10, 000 people? and it's like, how do you make it an initiative? How do you roll that out? and then, how do you create a strategy? and I don't mean like a five year plan because you're going to be wrong. okay, how do you go from business and then we talk about value creation, right? How do you take the Business strategy and break it down. And how do you infer the AI strategy from it? And we bring, an example there and built on the strategy model from Lefley and Martin. and then, and it was just pretty simple. So we were really showing how this can be done and how other organizations have done it, and then, talking about, one of the things is like aspirations, right? And what AI is supposed to do for your organization. if you start with automate our organization. Nobody wants to work with AI. If you start with, the most important thing is to help the people that work there. And it needs to be, genuine. then, people might actually be willing to work with you. So it's it's really important to have that strategy, to make it after you have a couple of POCs and after you have a couple of pilots to, to roll it out. And then, talking about project management and, waterfall doesn't really work. Scrum doesn't really work. So what works? So things in the middle. So how do you do iterative project management and keep it agile, but, have some type of planning and the better you, the further you get, the better you get. then we are, on the way to culture. How do you create an, an, a, culture that is actually, Embracing AI, right? And how can you help people to, grow in that culture? And we talk, we take the culture apart a little bit because it's a very intangible construct. So we talk about different, the hierarchy of culture from senior executives down to operations and how do you get all of those on board? And then, the last chapter, last part of the book, our three chapters about the technology, data, Data management and, talent. So when we talk about the capabilities that are required in order to create value, we have all three that are required to drive technology, data management, and different. different, skill sets, right? We always talk about data scientists and engineers, but there's way more to it. And, as we talked about the different, the five different big factors, right? They're usually not represented by those two, roles. So you have more roles that are required in order to be successful. So the skills can be, Taken by different roles or how you call them. It doesn't matter. So you need to represent the skills across the team And I think that's the important piece in order to create success. But yeah, so it's 350 pages So we are we wanted to keep it short Didn't work apologize for that. we apologize for every mistake we made in the book and everything that we might have missed and um, so we're looking forward to get feedback and You know, the people can reach out and please talk to us. So I have so many wonderful discussions about the book and what people think about it.
Don Finley:and since you brought that up, how can people get in touch with you? Are you on social media?
Michael Proksch:Oh, yeah.
Don Finley:the best path to Michael?
Michael Proksch:When we're, it's not just me, it's we have Nisha Pallival, managing vice president of Capital One. she's a co author of Willem Wieler, Dr. Willem CIO of Primatech in Canada. So we have, so please reach out on LinkedIn, right. Please reach out on. in the book, you have the opportunity to, there's a barcode, a website to the book where people can download more material. We can have discussions around it. so we're inviting everybody who wants to join the discussion to come and talk to us.
Don Finley:Dude, this is fantastic. I gotta say, thank you so much for bringing us. through this book. I like, there is, it's a topic that we're having every day, right? Or a discussion that is going on and ensuring that we're actually implementing this in the right places and understanding like what, that we have everything met, not just the technology, not just the data, but the culture, the people, the, the concern, the consideration for all parties in this is rather, just lovely. And like you've talked about it, it's a holistic approach. and it is based on. what we've done in the past and also how humans are part of this system. And really, like we build systems for humanity anyways. there isn't anything else that we're really looking at besides, the rest of the world as well. So I gotta say, what is the, what are the things that you're most excited for in the world of AI value creation And really just in general, whether it's personal, professional,
Michael Proksch:I think the most exciting part is when people talk to me about my job and. what I'm doing. I've been doing it for a while now. I like the creativity of it. and I think that's the more we use AI to do things that are and they leave us the space to be wise and create something unique. Because, and you really think about how many things we do every day that are just repetitive. you do because you have to, but don't really create anything there. There's no, nothing new, nothing special. Um, and that really fills my day a lot. but try to reduce it and I Try to focus on the more creative things that really create value. So I try to, yeah, I reduce the value, how do you call it? The detractors, more and more. that's one of the things I'm really excited to have AI for. especially when AI starts to mow my lawn and, and then, cleaning the kitchen, that would be cool. now we have a, iRobot that's, that's doing the, the floors, but, that's unfortunately not enough.
Don Finley:I know. I can't reach the countertops. That's the disappointing part. Yeah.
Michael Proksch:So someday I, the only thing is then what do we do with the time we get, right? Do we use it for something wisely and great value? It's just, use it to consume. it's everybody's decision to make, right?
Don Finley:That is a fascinating topic. And I think I want to have you back on to, to talk about possibly like the future after we get to that point, because. I know for myself, it's a mixture of the more that we've implemented AI, I'm taking a little bit more time off away. But at the same time, I also ended up creating new projects.
Michael Proksch:I always try to thrive towards, vacation time and getting bored. Can't get bored. I just, I always try to, Find something that I like and create and do things and fail. I failed so many times and I don't have a problem talking about it because I learned from it. I learned, from all the failures. That's why we have the book. that's, uh,
Don Finley:really fantastic.
Michael Proksch:I'm still struggling to embrace failure. It's live with it. let's say that.
Don Finley:But I think that's, and I'll wrap up on this because I think you're hitting on a key point, right? AI adoption and the reason why it doesn't work with waterfall and it doesn't really, that's that hybrid between the waterfall and the scrum aspect of, is you need that. iterative learning cycle that comes through from that failure and from that experience to see how you can actually adopt this technology. And it is new for us. It's a new toy. It's something great to be playing with And there's a lot of value there. But at the same time, if we don't create that cycle of being okay with creating an experiment that doesn't work out, then that's, not where the, that's where the learning happens. And so we need that. Yeah.
Michael Proksch:totally right, but there's a second piece to it, I think. And that's it needs to be an environment where failure is accepted. if there's huge risk to failure, of course you shouldn't play with it. Number one. And number two is you need to learn from the failures you make. So if there's no learning on there, that's maybe a future, part, or let's say chapter 15 of the book is like organizational learning, where we are, how do we learn from the mistakes we make? If we don't talk about mistakes, if we hide the mistakes, if we're only talking about the successes, we can't win and that's the, that's the other side. So how can we. How can we learn from things that we made, that we did wrong? Um, and we need to be honest about it to ourselves and to others. maybe more to ourselves than to others because
Don Finley:Oh, Michael, that's a great point. Again, really appreciate it. but so yeah, I think we got to create cultures of learning and cultures of, understanding that hey, we're going to fall down, but We'll pick you back up. And this is a time that, we look at everything that we can possibly look at. So really great having you on. I like, I've always enjoyed our conversations and so I'm just glad that we get to share one of them, with the. world. And again, thank you so much.
Michael Proksch:Cool. Thank you so much.
Don Finley:Thank you for tuning into The Human Code, sponsored by FINdustries, where we harness AI to elevate your business. By improving operational efficiency and accelerating growth, we turn opportunities into reality. Let FINdustries be your guide to AI mastery, making success inevitable. Explore how at FINdustries. co.
