[ BETTER TECH LEADERSHIP ]

Mirko Kämpf: Teaching, Trust, and Technology - A Journey to AI-Driven Leadership

[ THE SPEAKERS ]

Meet our hosts & guests

Leszek Knoll
CEO, BRAINHUB

Over the last decade, Leszek has developed several successful businesses, among them a software development agency that supports Fortune 500 companies. With the challenges a growing business brings, he observed that stepping out of a tech role into a leadership one brings the need for a different approach. As a host of the Better Tech Leadership podcast, Leszek is focused on bridging the gap between tech and people skills.

Mirko Kämpf
CTO & Chief Architect

Dr. Mirko Kämpf is an expert in event data processing and time series management at scale, driving innovation in data-driven business architectures. He specializes in stream data management, knowledge graphs, machine learning, and GPU-based processing. A mentor and advisor, Mirko empowers organizations to harness AI and scalable data solutions for real-time insights and decision-making.

Transcript

This transcription is AI-generated and may contain errors or inaccuracies.

Leszek

My name is Leszek and I will be talking with Mirko Kämpf about fostering innovation through leadership and adapting to technological change while inspiring growth and collaboration. So Mirko, could you start by giving us a brief overview of her career journey? What led you to, you know, from one role to another, how your career career evolved throughout the years?

Mirko Kämpf

Over the years, my career evolved in multiple different ways, non traditional step in between. So I started as a craftsman, stepped into a more industry like context, the military, but more the technical part of military. So I was involved in complex technical systems and from craftsmanship to industry level work, there was a big jump, but there were so many similarities and this was making me very interested in learning about the background, what's going on behind the scenes. This is what drove me towards the next level. And then I said I need to study and understand deeper what's going on. And so I had to study physics because in physics you learn how to get an understanding, how to describe systems, how you make them ready for being able to talk about it.

That's the point. If you can't give things names and can't talk about it, it's hard to manage things. And this is something I learned in physics. And then I said, okay, this is just my toolbox. I never intended to become a researcher, but I was intentionally going into the field to fill up my toolbox and this worked out. And then I jumped into the industry again in the big data industry at this point in time, then I could apply the methodologies from physics, from craftsmanship into the day to day work. And so being creative and also in a leadership position within a project, not as a team lead, but as a project responsible person, I could see how valuable all this is if you get an idea and can somehow translate it.

And this is what teachers do. And this was my thing, this is where all things come together. So by learning fast, translating things into a language others understand, I think this makes good trainers, good teachers and also good leaders. And so I never escaped from that domain or from this field of influences.

Leszek

Do you consider it just, you know, a tool for actually describing and understanding reality or basically the tool set that physics provide because it's not specific to, you know, the thermodynamics or anything is. I, I think it's more what you are saying is actually the physics is the way of thinking about reality that could be very used, useful in other domains. Is that, is that the, that the message?

Mirko Kämpf

Yes, that's the message. In my specific field we called it socioeconomics and social Physics, so the connection between social actions between people, society as a complex or larger structure, and the interaction within these systems on different levels. All this is stuff which is modeled in different disciplines within physics, but not only using a field like thermodynamics to describe something, but also the methodology, how to learn how to work with thermodynamics. Also this is a very important tool for, yeah, for getting ready not the physics results, but the way how physicists work. Then it was even more informative to me. Not the, the expertise in becoming the expert in thermodynamics. That's not my thing.

That's something which is a result of the work. But when you use the tools and when you are able to use this scientific approach, no matter what kind of science it is, the scientific approach and the scientific thinking is so close to craftsmanship, the creative element in it is so important. And this was really driving me and this is the moment where I said, no, you haven't to be a genius. To be a genius may be something how others recognize you, how you look to them. But to be a genius is not something you are or you are not. Maybe it's just the way how others observe you. The question is, what can I do to do things right, to do no harm, to be fast, to be on point?

And so yeah, this is very important. If your tools are sharp, you are fast.

Leszek

Typically this probably this toolset is very useful also in the AI space, I suppose as much as in any other space. But you've been working on a framework for AI readiness and I like to unpack this topic. This sounds, you know, incredibly relevant for organizations for this, the, the current times as well. Could you walk us through what AI readiness means and how the framework could help?

Mirko Kämpf

Yeah, AI readiness, what does it mean to be ready for AI means either as a person or as an organization. You can be ready. But ideally, and I think it's a strong requirement, you must be ready on both levels. As a company, as an organization, you must provide an environment where people can learn and experiment in a safe way to get AI ready. But on the opposite side, if people are not ready to start the experimentation, nobody will bring this organization to the AI ready level. It's a chicken egg problem. And so different people argue differently.

Some say you do this as a top down. If the management has no AI strategy, we are all lost. Yes, of course a company needs an AI strategy for the future, but without having people who are skilled to do the stuff, the best strategy doesn't help. And that's the point you need on both levels, different things which define AI readiness. In my framework, it means if the management knows what AI aspects are relevant, they are AI ready. And then on the other layers, on the middle management or on the layer where the work is done, the question is now, do we have the right tools or the skills to use the tools? Ideally, both.

And the reality is, even if there is a nice tool on the market, a product, a service, typically it doesn't fit directly into your environment. It must be adopted, it must be integrated. And AI readiness now means to understand how to implement a service, a product, a tool, a methodology from outside, so that our company can drive things faster using this tool support.

That's the whole thing. And sometimes you need little adjustments, a little bit of learning, of adoption. Sometimes you should really build a new factory besides the old one, because the old one doesn't fit. It depends. And figuring out these details is part of the AI readiness procedure, this analysis procedure, to understand where is the potential and what do you get from it when you start using AI? We have seen places that people couldn't say a word. They were so impressed by a result, by the first result.

And then we said, wow, this is cool, let's continue. And after doing a little thing three, four times, some prompting with ChatGPT, they are still amazed, but then they fail in adopting it. On a recurring level, this means you are not AI ready, you are ready to play with things. Yes, that's cool. But bringing this new idea, this new capabilities into your processes and adjusting your processes, this is what AI readiness means. @ the end of the day, the.

Leszek

Example that you're referring to actually speaks that it's pretty clear. How is it relevant what you can get out of it, to your points, to your previous points, but the rollout of this transformation is either unclear or just fails. And this comes down with my understanding to the tools, skills, or basically environment on the bottom level that actually is not ready for that.

Mirko Kämpf

Yes, it's a mix of all three things, the tools and the skills available to the bottom level, then definitely one part of it. But I do not think about blaming the one side or the other.

It's about identifying gaps. And the question is, who should inspire the leadership, if not the people in the bottom?

Leszek

Yeah, absolutely, absolutely.

Mirko Kämpf

Who should inspire the people in the bottom layer to do this additional stuff to give them freedom for exploration, for testing, to figure out what can be done better? This must come from management. If people have no time to study, to experiment, to learn, to teach each other, to do Things faster if this time is not given to them. They could be, yeah, they could have the idea, but management must support it. And therefore they must also trust that something comes out of the box. And here's the problem. If people in the bottom level say, why should I be x percent more efficient?

What am I doing with the time I get back? I have nothing. On a personal level, this is what people typically say. And management must now figure out how to deal with this.

Leszek

All right, all right. So what would be like the first initial steps, how to, how to tackle, how to actually build that scalable strategy to, you know, to be actually ready? How would you tackle this?

Mirko Kämpf

We tackle it on both sides. We talk with the management, identify on the business layer the potential. Where is the data? What kind of data can be used? What is the behavior which should be changed or automated? Where is the cognitive overload which must be or which should be reduced by using AI? This is the trigger question.

Do you have a topic where a cognitive overload has been experienced? Is there something you would like to do? Maybe processing tons of transcripts or video data or interpreting photos? This is an example for a cognitive load which can't be handled without AI. With AI, you could do this easily if you know how. And then we dive in and say, okay, on the business layer, we do not talk about ChatGPT day in, day out. It's on the bigger picture.

How can your business be transformed to do things which you can't do today? That's the question on the business layer. And at the same point in time, it's all about giving people on the bottom the tools to be fluent in the new tech. So what does it mean? You should be fluent to transcribe an audio recording or a video recording? Because this is something which is for free, more or less. Not in terms of the services for free, but you have not to do a lot if you know how.

It's like learning how to use paper and pen. There is paper, there is pen everywhere. But if you don't know how to write, okay, most people are aware of how to write, but still they can't write an essay or a poem. We know how to concatenate letters and words and build sentences. That's the level where we are very fluent in society as well. But the next level is missing. And we shouldn't talk about essays before we know the Alphabet.

And that's the starting point. Give people the Alphabet, the pen, the paper, and then bring them together to start writing letters to each other. And then as a Result of being active. So writing letters is just an example. But let's assume we write letters to each other. Maybe a third person starts collecting the letters and makes a book out of it. We would never have made a book.

But the other person from an external view can now also start collecting. And this is a key ingredient. You must work on different layers of abstraction and let people do what they need to do. And then others could group and organize and aggregate and arrange things which are out of scope of the people in the one layer. And this is my middle management Johnson. As a normal team lead, I shouldn't compete with my team members who is writing the cool articles. No, I should aggregate the knowledge and figure out who needs new knowledge.

And this is where leadership and training come together as an element of the onboarding of AI onboarding, you need to lead and train at the same moment. You can't separate this. At least this is my understanding and our approach.

Leszek

Okay, okay. When there's, there are probably some trade offs here and a lot of them probably, and I want to talk about one of them is actually, or maybe, or maybe let's forget about the trade offs and talk about. There are probably situations where it actually makes sense to follow the AI trend and sometimes it probably doesn't. I think the framework actually sort of clearly helps in navigating this decision, especially the things that you said, for example, knowing how relevant things can get, I mean, knowing how relevant aspects of AI are or what is it in for you in that. But could you elaborate on, you know, the decisions whether to actually use the AI or not and how to, how to make this decision. This is even, can be even broader discussion because AI is one thing, but also in terms of building, for example, software. There are also, you know, the sort of, this fine line between whether you use, for example, algorithmic approach or you know, machine learning, data engineering sort of approach.

It also, this thin line is sort of moving in one direction and it's probably going to settle at some, you know, balanced place. So this sort of decision is highly relevant on the organization level, but also on the implementation level. And I'd like to ask for your thoughts on that matter.

Mirko Kämpf

I can use this analogy with the writing and with the book. Again, not everybody who learns writing will end up in publishing a book for sure not. And this is the same with using AI. If the published book is the product of a company which gets out of the company to the customer base, the question is who is involved in building this product, this result. And if AI helps me to Speed up my learning experience to allow me. If it allows me to digest more scientific papers or to research through more books to identify a missing puzzle, then I have a huge value from using AI. If I know how you know, it's again like writing with a pen or a pencil, or with a brush.

It is a technique I need to know. And depending on the techniques I know, I can be faster or of a higher quality, more on point. And all these things, these are the criteria. It's not a yes or no, it's a question. What ingredients are needed? And where is the AI technology supporting us? Sometimes it's taking away routine tasks from me as a human being.

It allows me to automate things. And we have seen this in the industry. Automation of melding or of drilling. All this stuff can be automated. Highly sophisticated technology, very sensible, but also robust in between. Cool stuff we could not imagine for 50 years ago. And now we have a transition with AI.

The same automation is in front of us, but for cognitive tasks. Was it this skill of the hand to move the hand in the right angle, in the right speed? This is a craftsman skill. And now we automate the thinking skill. I do not say we automate thinking, but the cognitive skills of the human. Recognizing things, classifying things. This is what we can do automatic.

And if we learn where such an automatic classification helps me in delivering my product, then I can answer the question. It's not yes or no in the beginning, it will emerge. This yes or no is a result of an analysis along the value chain. And the question is, do I need an AI inside my product, or do I need an AI on every work desk, for every worker? But the final will, the final result could be composed entirely manually by the people, if the policy allows requires this. And that's the point, there is no question. AI is not allowed for us.

The question must be, where is it forbidden? Where can it be allowed? And if the human in between stays in the center of the whole thing? We are human society and we should then not be anxious or depressed by being replaced by AI. The question is, can AI become our body, our best friend, our supporter, and every person who does some stuff in the workshop or the machines. Tools are all automatic to some extent. For our hands, it's already common.

And the next thing is for senses, for our eyes, for our ears, for our spirit speaking. Or they are not technical supporters, technical assistance for individual skills. And I think that's the key for the moment, where the society can do a huge thing and then as a consequence, a whole New layer of applications will emerge.

Something I can't imagine today.

Leszek

That's a nice ways to. That's a nice way to. And this block which is actually it's really hard to imagine what's coming next. That resonates with me very much. Now I'd like to talk about the concept of learning by teaching. I'm a big fan of this approach actually. I mean naturally this came to.

I sort of. I'm a big fan of this approach. I've learned it not by reading books, but actually by exploring but by trying to teach somebody and, and actually drive my growth in a way that was. Come really hard to compare with anything else. And at least that's my experience. And I'd like to get your thoughts on that. Whether, how do you use it, whether actually it's part of your approach on daily basis.

How do you. But basically I'd like to get your thoughts on learning by teaching.

Mirko Kämpf

Yeah, learning by teaching was an inspiration. I got it when I was learning Spanish from a woman. She spoke Portuguese and she said I give a Spanish class to improve my Spanish again and that's my way of being active. Learning by teaching. This was so impressive and I was inspired by this thought. And I figured later out that teaching can be one of two things. Either there is something known in the world and others don't know it.

And then somebody has to share that knowledge. It exists already. And maybe textbooks do exist, trainings, courses do exist and somebody enforces others to learn. This is what people usually say, this is teaching. It's something which is well understood and known. But then there's the research world. In research you figure something out which is maybe a new insight. And in order to have partners to work with, you need to share your knowledge.

It's also a kind of teaching because you know it. Now maybe you are the first person who knows something new. And in order to spread the word, you have to share these things. And it's a kind of teaching. This means in order to go deeper in your own domain at the same point in time when you want to go deeper, understand deeper, you need to share the things. Why? Because you can't go deeper without a sparing partner.

You need somebody who starts the ping pong with you, who reflects with you. And this way of I tell you something with the goal to, let's say, teach you some new idea, you absorb it and then you do the same thing. You go back to me. Now we are in this inter exchange mode and this means we both drive the new insights we both link the new insights together and this means we can grow together. Although I initiate it, maybe, maybe I'm in front of you by one week, but I could not go to week five with success without having this sparing partner. And this is why I say learning by teaching. You need to start in a new field, typically not very alone.

But if you have a partner, a sparing partner, or a group around you, then you have more energy and the whole flow is more intense. And this is what works more or less in every domain. Even if the world has understood how to do things, maybe there is a group of people somewhere which has no access to resources, but can explore the surroundings on their own. They would be able to learn the things even without the book self exploration. And this is so interesting because we all say data analytics, self service, what's the difference? Here I go into a data lake, there's lots of data, I have no clue what I can found. And now I have to explore the data lake like I would explore the forest for building my camp for the evening.

And this is the metaphor. Nobody is there who tells me where to start in the data lake to dig for the really important things. And this is the same in real life. And if we look back in such human patterns, how did I find out where to do my camp in the forest? Was there a master who dictated the rules I had to follow? Ah, in driver school it was like this. First there are the rules and then I follow and then I drive safely.

That's a different world. Having fun in the forest is different. And so I think this fundamental elements allow us to understand better how people behave and why and what to do next. It's always the question, what's the next step? Is there a book which tells me what step, the next step, or do I have a methodology inside of me to do the next step? And this brings me to this learning by teaching. Because it is coupled, we can't separate it.

And that's the chicken egg problem. If we see it in a comprehensive picture, the chicken egg problem is not really a problem, it's just a thing which needs to be handled. And so many people are distracted or afraid of this chicken egg problem, they step away. Because how should I do it? The question is, don't worry, start somewhere and then go do it here. And then try to move until you make it somehow rolling. This is the big thing.

So waiting for the teacher is not the best idea if you want to learn fast.

Leszek

Exactly. I think, I mean, there's one part that actually I mean, maybe a side comment here is I think people very often get paralyzed with the chicken neck problem. And that's, that's one of the things I learned off the way. Yeah. And just start somewhere nice. Nicely put. Your approach to learning, you know, ties closely to leadership.

And as you look back on your career, what are some of the, you know, leadership lessons you've learned that shape your, shape how you lead, shape how you guide others? Probably lessons learned, approach thing is one of the things, physics is one of them. But I, I'd like you to cover, you know, your main leadership lessons learned.

Mirko Kämpf

Yeah, the main leadership lesson for me is take responsibility to stop thinking followed by the doing. And as long as you do not step on the foot of other people, this is something you must have very clear in your mind. Not to the advantage for you with a disadvantage to others. But finding a gap and filling a gap with something new is something you should just do until you reach the boundaries without asking. Because if you ask others to say it's okay to do this, then you are not leading by asking is it okay to go into this gap? You stop leading. You will be led by the person who decides for you the yes or the no.

And I think that's an important thing. Leadership means to be straight, to be clear and to go forward. Because otherwise people can't follow you. Leadership is not being behind the group and pushing them and pressing them. It's about being in front and go forward, explore the room, come out of it and tell others your experience, share the experience and make them interested in also going into the room instead of pressing the people into the room and telling them what to do or not to do. So this is how self exploration is the baseline for being able to lead. Because if you can tell them the nice experience and warning and make safer for them to go in as well, then this is leadership for me.

It's like exploring the forest and then coming back and telling what definitely not to do to avoid unnecessary losses. What about making the self experience faster and safer for people by avoiding redoing the same mistakes?

Leszek

But I think there's also. You cannot go fully into protecting the people. I guess, I mean it's not protecting.

Mirko Kämpf

Absolutely not. They must be self responsible for what they are doing. But if you have an obvious hint, please do not eat this herb, it will kill you. It smells so nice, but it will kill you.

I've seen it. There was the fox and the deer dying. I saw them dying. Then this warning not to share this warning. This would be A failure if it happens information. But if people want to experiment with it, maybe they start collecting it, cook it, make something out of it and it works.

Leszek

Exactly.

Mirko Kämpf

Openness must be there. And this is trust. It's all about do I trust my next neighbors? The building trust in your next neighbor neighborhood is in very essential element.

Leszek

I think there's also one element in that which is really important. If, if. Maybe that's my way of thinking, but maybe you think differently. But it's also really important who you take on this journey to this forest. I mean one of the things is, you know, shaping how people think, how people act, influencing you. But also, you know, this, I don't want to say selection because it's a bad word, but actually, you know, paying attention, who's on the trip, whether these are individuals who you want to take on the trip, who will, you know, operate as under these, this leadership. Maybe not take full, I mean that will basically take full advantage of this approach.

Because that's probably not for everyone, I guess.

Mirko Kämpf

Yes, I absolutely agree with this. And this is also my observation. There's a, there is a continuum from such self responsible leadership to followership which also has a part of leadership for the next level down to people who will follow and be very supportive and very loyal. But they are not able to lead anything, not even themselves. And those people exist. And you need as a leader to figure out on what level you can deal with who. For some people, a very straight procedure, a very clear description of the procedure is essential.

And others say, go away with your procedure, tell me the goal and I deliver. Anything else would kill me. Different people, the different approach of different people can now be utilized. You can say, okay, a new task, which has never done before, must be done by someone who drives themselves to the solution. And if this person documents the thing, I can take this as a blueprint, let it repeat by somebody in between. And then finally we do a standard operations procedure to give it as a manual to everybody. And this is the same thing how we introduce AI in a company.

Some will explore the next, will validate and adapt it to the company's needs to standardize. And then we can say, okay, this is now proved, it's working, it's safe and everybody now can use it. This is also a multilevel leadership thing. You can't say I tell you how to do things always and sometimes you must do that and to figure out where, which intensity of command is needed. This is also in leadership thing.

Leszek

Yeah, yeah. I mean coming back to the forest exploration thing. You need people who will basically explore the area, but also you need somebody who will chop the wood for the fire. And he's perfect at that, right?

Mirko Kämpf

Yep, absolutely.

Leszek

Mirko, thank you very much for today. It was really some really valuable insights. Thank you also for delivering the passionate and engaged way. I really enjoyed our conversation. Thank you very much. It was a pleasure.

Mirko Kämpf

Thank you also for inviting me. It was an interesting discussion. I love it. I love the topics and also your approach to share these ideas with the community. So I'm thankful for being a part of this.

And yeah, enjoy your day. Have a good one.

Leszek

Follow Matt and Leszek on LinkedIn.

Explore similar episodes

Martin Miller: The Art and Science of Building a Successful Startup

In this episode, Matt and Martin Miller discuss bridging the gap between technical teams and non-technical founders in startups. They talk about the importance of collaboration, communication, and realistic expectations. Martin also shares his perspective on AI, machine learning, and how companies can maximize their return on investment in AI/ML initiatives.

listen now
Sonny Patel: From Developer to Leader - A Journey in Product Management

In this episode, Matt interviews Sonny Patel about avoiding burnout and driving technological innovation in traditional industries. Sonny Patel shares her experiences at Microsoft and Amazon, and how it shaped her views on product management and team leadership. Sonny also discuss the impact of AI and how it changed the way she works.

listen now
Michael Frembs: Mastering Team Structures - Insights into Platform and Streamlined Teams

In this episode, Leszek interviews Michael Frembs about structuring and leading tech teams. Michael shares his career journey, highlighting key milestones and leadership lessons. He offers insight on how to build effective teams, foster a culture of trust and collaboration, and the importance of adapting team structures to meet evolving needs.

listen now
Robbert van Os: From Startup to Scale-up - Key Auditing and Leadership Strategies

In this episode, Matt interviews Robbert van Os about conducting tech audits. Robbert shares insights on common challenges startups face, such as limited resources and the pressure to adopt popular methodologies. He stresses the importance of leadership in maintaining an open mindset towards constructive feedback and adapting to the evolving tech landscape. Robbert shares practical strategies and real-world examples illustrating how companies can navigate these challenges effectively while maintaining their competitive edge.

listen now