#566-#577 Who Leads When AI Thinks: The Complete 12-Part Conversation With Dr. Adrian Wolfberg
Who Leads When AI Thinks: The Complete 12-Part Conversation With Dr. Adrian Wolfberg
Over twelve consecutive episodes of The Leadership Podcast, host Niels Brabandt sat down with Dr. Adrian Wolfberg, author of Who Leads When AI Thinks, for what became one of the most comprehensive conversations available anywhere on what artificial intelligence actually demands of leaders. Not technologists. Not data scientists. Leaders. This article draws the full arc together for business decision-makers who want the complete picture in one place, before reading the entire series transcript below.
It Starts With a Dilemma, Not a Tool
The series opens with a deceptively simple observation: AI is extraordinarily fast, and that speed creates a genuine leadership dilemma, because someone still has to interpret, judge, and take responsibility for whatever it produces. Wolfberg is emphatic that Who Leads When AI Thinks is a leadership book, not a technology book, because for the first time in history, human and machine consciousness effectively share the same working space, with no historical precedent to draw on. The leader's job starts before the tool is even opened, with the question of what problem is actually being solved, and whether AI is even the right approach to it. Poor thinking, magnified by AI, produces sophisticated mistakes rather than fewer of them, which is why questions now matter more than answers, a genuine inversion of decades of management practice built around leaders having the answer.
Leadership Follows the Problem, Not the Technology
Episode two tackles the fear many leaders carry privately: that their teams will trust AI more than they trust them. Wolfberg's answer reframes the question entirely. Leadership is not decided by which is more capable, it is decided by the nature of the problem, assessed across familiarity, complexity, and ethical weight, plus a hidden fourth dimension, time, since who leads today is rarely who leads tomorrow as a problem evolves. He is reluctant to concede that any scenario justifies zero human involvement, even in near-automated systems such as nuclear reactor monitoring, because human interpretation remains necessary even at a ratio of ninety eight percent AI to two percent human.
Four Ways Humans and AI Actually Collaborate
With the leadership question settled, episode three turns practical. Wolfberg sets out four collaboration modes: human only, for situations dominated by ethics, trust, or confidentiality; AI-enhanced human, the most common mode, where a person prompts, interprets, and remains the ultimate decision-maker; dual engagement, where two people each run separate AI conversations and must then reconcile potentially conflicting interpretations; and AI only, appropriate for routine, low risk, stable situations, but which Wolfberg insists must always be a deliberate design choice, never a default, and should always be treated as a warning sign worth examining.
Every Framework Assumes the Problem Stays the Same. It Never Does.
Episode four introduces adaptation as a distinct discipline from flexibility. Adaptation means changing your understanding of reality first, which then changes decisions and actions. The starting point is framing, deciding what problem you are solving and what success looks like, and the same facts, declining sales, for instance, can be framed as a marketing problem, a product problem, or a customer experience problem, each pointing to a different solution. Perhaps the most counterintuitive claim in the entire series follows directly from this: flawed framing should be treated as a welcome, expected part of decision-making, not a professional embarrassment, and disciplined inconsistency, changing your mind for a defensible reason, is a sign of intelligence, not weakness.
Diagnosing the Problem Before Reaching for a Solution
Episodes five and six deepen the diagnostic toolkit. Wolfberg's three-dimensional cube, familiarity, complexity, and wickedness, assessed one dimension at a time and then combined, lets leaders name a problem precisely rather than vaguely. Most leadership failures, he argues, begin with misidentifying the problem itself. Episode six adds decomposition, the discipline of breaking a large problem into meaningful, understandable pieces without losing sight of the whole, illustrated through the famous failures of New Coke and the opioid crisis, both cases where a narrow decomposition of the problem produced consequences the narrower frame never anticipated. There is no single correct way to decompose a problem, by time, stakeholder, function, cause and effect, or geography, and the right strategy often changes as conditions do.
What AI Cannot Replace
Episode seven confronts the question underneath the entire series: if AI becomes capable enough, what is left for human thinking? Wolfberg's answer is that AI does not eliminate human thinking, it changes which kind of thinking creates value. Routine analysis declines in value. Judgement, creativity, and interpretation rise in value, because critical thinking, evaluating evidence and questioning assumptions, and creative thinking, generating possibilities and reframing problems, remain distinctly human capacities that AI cannot reliably perform.
Three New Leadership Competencies
Episodes eight through ten introduce a trio of competencies Wolfberg argues organisations have never had to formally develop before. Adaptive capacity is the ability to absorb new information and update understanding without becoming disoriented, diagnosed through two questions: how overwhelmed am I by volume, and how understandable is what I am receiving. Mental acuity turns raw information into genuine understanding, distinguishing signal from noise and, crucially, making time for reflection, the most neglected leadership habit, because slowing down is what allows assumptions to be examined properly. Boundary crossing addresses a counterintuitive risk: AI can connect information across disciplines, but only leaders can connect the people, and without deliberate boundary crossing, AI can actually widen organisational silos rather than closing them, as engineering, marketing, and legal each receive excellent answers from their own narrow AI conversations that drift further apart over time.
From Framework to Practice
The series closes with implementation. Episode eleven names three organisational failure patterns, framing errors, where the wrong problem gets solved efficiently, alignment errors, where teams agree on the problem but not its priority or ownership, and suppression errors, where dissent quietly disappears and every recommendation is accepted without genuine challenge. Most AI initiatives fail, Wolfberg argues, because organisations focus on deploying the technology rather than changing the behaviour around it. The finale distils the entire series into a practical checklist: make reframing a habit rather than a crisis response, build adaptation into the organisation deliberately, invest in the five competencies covered across the series as strategic assets, protect human thinking by refusing to outsource framing and judgement, and build a genuine culture of questioning.
The Enduring Point
Across twelve episodes, one message recurs in different forms: AI does not simply change how organisations use technology, it changes how leaders need to think. Technology will keep evolving. The need to frame problems clearly, exercise sound judgement, integrate competing perspectives, and adapt to a changing reality will not, because those remain enduring leadership responsibilities, whatever tool happens to sit on the desk. The full, unedited transcript of all twelve conversations follows below, for readers who want to go deeper into any single episode.
Niels Brabandt
---
More on this topic in this week's videocast and podcast with Niels Brabandt: Videocast / Apple Podcasts / Spotify
For the videocast's and podcast's transcript, read below this article.
Is excellent leadership important to you?
Let's have a chat: NB@NB-Networks.com
Contact: Niels Brabandt on LinkedIn - https://www.linkedin.com/in/nielsbrabandt
Website: www.NB-Networks.biz
Podcast and Videocast Transcripts: The Complete 12-Part Series
Episode 1 of 12 (E_566): Speed Without Judgment: The Leader's Real Dilemma With AI
Niels Brabandt EMBA MBA MSc
We now face the age of AI, and of course we see a lot of challenges here. And I'm very happy that we have an expert with us here today for quite a series of episodes, 12 in total, and we are going to talk about what is the leader's dilemma when it comes to AI first. Hello and welcome back by popular demand, Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you for having me, Niels.
Niels Brabandt EMBA MBA MSc
Thank you very much for taking the time. Let's get straight into that. What's, in your opinion, the core leader's dilemma when it comes to AI? What is the challenge that most leaders face?
Dr. Adrian Wolfberg
Well, there's a dilemma, and that dilemma is: on the one hand, AI is very quick. It can process information quickly, it can—
it can create pattern recognitions, it can— it can take a lot of information very quickly and provide an output. But on the other hand, there's— there's this need for humans to interpret what is the output of AI. And if— if humans don't take responsibility for
interpreting and judging what comes out of that, then humans will default into letting AI become the sole or the primary answer to questions, or to frame— to come to a solution, or so forth. So that's the dilemma. On one hand you have speed, and on the other hand you have human content that's essential to not give up.
Niels Brabandt EMBA MBA MSc
Mm. When we now look at AI, most people say, "Hey, I just tell people, maybe you take a look into Claude or ChatGPT or Perplexity or what it is." You wrote the book Who Leads When AI Thinks, and you call it a leadership book, not a technology book. So why would you say it's a leadership book, not a technology book in particular?
Dr. Adrian Wolfberg
Yeah.
Well, I mean, every technological revolution ultimately becomes a leadership challenge. And with AI, it's more than a tool. For the first time in human history, consciousness now lives in a shared space with machine and human. So there's a shared consciousness that's never been experienced before.
So if you look at prior tools, these are technologies where we knew the inputs, we knew the outputs, we had predictability, we had knowledge of the process, of how the technology was used. With AI, there's nothing like that. So the question is, what should leaders do differently because AI exists?
And so this leader's role begins before even touching AI. Leaders should ask, what problems are we trying to solve? Do we really understand the problem? Is AI even the right approach, and for what parts of the problem? Which means where should AI, and what type of AI, should be involved?
The risk is that poorly framed problems, from a human perspective, will produce impressive-looking but disappointing results. And AI has almost never fixed poor thinking. So poor thinking, to begin with, yields poor outputs. So it doesn't replace problem framing, but it magnifies and reinforces whatever you start with.
Niels Brabandt EMBA MBA MSc
Yeah.
Dr. Adrian Wolfberg
So this means that questions are more important than answers. Whereas in the past, leaders have traditionally been rewarded for having answers. In the AI era, those who are rewarded are those who ask the better questions. You know, what data gets collected? What assumptions matter? What AI is asked to optimize? What success looks like? So better questions produce better AI. Poor questions produce sophisticated mistakes. So yeah, so—
Niels Brabandt EMBA MBA MSc
Sophisticated mistakes is also a nice term, yes.
Dr. Adrian Wolfberg
Yeah, and we know from the history of smartphone usage is an indicator of what we're seeing already with AI. So smartphones, you know, over the last 20, 25 years, have become our external memory. People remember where information is rather than the information itself.
Niels Brabandt EMBA MBA MSc
Yeah.
Dr. Adrian Wolfberg
And these constant notifications reduce sustained attention. So anything more than 30 seconds or 60 seconds sometimes becomes a challenge. And so we're increasingly outsourcing remembering, and AI now encourages us to outsource reasoning as well. So this is, I think, a critical challenge.
Smartphones changed how we remember. AI may change. I think it will change how we think. So what we know already about AI, large language models like ChatGPT, Claude, etc., is they're great at pattern recognition, great at summarizing, great at generating options. But they're unreliable at judging context. Context is critically important.
They cannot determine priorities, cannot determine what's an acceptable risk, and it doesn't own the consequences. So, you know, we— we've heard stories, you know, it's documented that the greatest AI risk is hallucinations from outputs from large language models. And I would say that's not— that's not the greatest risk.
Yeah, it is a risk, but it's that humans become intellectually passive. That, to me, is the greatest risk by defaulting to AI. And I think we're already seeing that in the early research that's being done in the last few years on university students and adults in the workplace.
Niels Brabandt EMBA MBA MSc
Absolutely. Yes?
Dr. Adrian Wolfberg
Yeah. So, you know, it's— what's happening is that we just accept answers that may or may not be reliable. And we're stopping asking questions too early, and we're confusing the confidence that AI generates with understanding. And so in the era of AI, you know, critical and creative thinking become more valuable. Routine thinking declines in value because AI can handle that.
Niels Brabandt EMBA MBA MSc
Yeah.
Dr. Adrian Wolfberg
But the need for judgment increases. So what I call framing increases, creativity increases, the need for interpretation increases, the need for adaptability increases. There are no checklists, unfortunately.
Niels Brabandt EMBA MBA MSc
That will be my question now, because many leaders will say, "Hey, Dr. Wolfberg, can you please give me a checklist with questions to ask in— in— in my team so I can just go through the checklist?"
So when there's no checklist, as the final question for this part of the interview, how can people know which questions to ask, especially when they say, "I'm not very tech-savvy, I'm not very senior with AI, how to know which question to ask when I'm on the junior side as a leader myself?"
Dr. Adrian Wolfberg
Yeah. So in my book, Who Leads When AI Thinks, I provide frameworks, and I actually provide specific questions. But these are leadership questions. They're not technology questions.
They're questions like, who should lead? How should people in AI work together? How does the work flow? What strategy may guide me in this human-AI collaboration? How should I adapt when conditions change? How do I recognize when we're drifting from reality?
So these are the kind of questions, more specifically, that I include in the book, and I provide these in conjunction with frameworks along the way to help people navigate this shared consciousness, this dilemma between human and AI collaboration.
Niels Brabandt EMBA MBA MSc
I think these are the perfect final words. And of course, now the question might be, who leads human-AI collaboration? So who is the real leader here? And that is going to be the topic of our next episode. So stay tuned, and for now, Adrian Wolfberg, thank you very much for your time.
Dr. Adrian Wolfberg
You're welcome.
Episode 2 of 12 (E_567): Who Is Actually in Charge? Rethinking Leadership in the Age of AI
Niels Brabandt EMBA MBA MSc
AI is here, and AI is here to stay. The question is: who is leading, who leads the human-AI collaboration, and we have an expert again with us here today for episode number 2. Hello and welcome back, Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you for having me, Niels. Great to be back.
Niels Brabandt EMBA MBA MSc
Thank you very much to be back, and thank you very much for taking the time again. We get straight into it.
Of course, people now say, look, AI is there and AI is here to stay. The question is: who is now in the lead?
Because when I prepared for these episodes, quite a number of leaders said, I fear a bit, that suddenly my people think that they can rely more on AI than on me, or what will be with my leadership role in 5 years, or 10 years, or even 2 years' time. So I pass the question straight to you: who leads human-AI collaboration?
Dr. Adrian Wolfberg
Yeah, no, that's a great introduction. And if I don't go into all of that here, we'll do it in later episodes as well. The big question is, you know, what is the role of leaders in the age of AI? Well, let me just recap what we talked about in episode 1, the previous episode. We asked why AI creates a leadership dilemma, and we answered that question. In this episode, I'm going to talk about how leaders navigate this dilemma.
So most leaders would immediately ask, should AI lead or should humans lead? That's a natural question. It's a natural questions that those working at the edge, working with the systems themselves, would ask. But I'm saying, I'm proposing that before deciding who leads, we must understand the problem. So what I'm saying, what I'm proposing, is that leadership is not determined by the technology. It's determined by the characteristics and the nature of the problem itself.
So, you know, what is that? Well, there's a number of perspectives that we can look at in understanding what the problem is. One dimension is, to what degree is the problem routine versus novel, or something in between? So routine means we've seen this problem before, we know how it works, we understand the inputs, we understand the outputs. On the other end of the spectrum, something is novel. Existing approaches are unreliable because this is something totally new. We haven't experienced it before. In order to experience it, we have to learn while we're actually experiencing it in order to solve it.
There's another aspect to this of understanding the nature of the problem, and that is, where on the spectrum is it between a simple and a complex problem? So a simple problem is where the cause and effect are clearly established. Relationships between people and variables are stable. There's few interactions. Complex problem, there's many interacting variables, and small changes can cascade in creating unexpected effects. And there's a third dimension, whether something is wicked or not wicked. Wicked involves an ethical or a value-based issue where there's disagreement on the problem itself and there's no understanding of what the solution should even be. On the other end of the spectrum, there's something that's less wicked or not wicked at all. It's little or no ethical or value content.
So those are just kind of an initial perspective that one should take about understanding the nature of the problem. But there's another layer above and beyond that, and that is whether we understand this problem from an internal perspective. So this means, to what degree is our experience relevant? What can we gather from the organizational memory dealing with a problem? What kind of expertise does the organization have dealing with such a problem? And different organizations will naturally answer these questions differently. The other is externally. So it's about the world around us, not about us, not about us itself. So the world has a great role in determining uncertainty and interactions and unpredictability.
And then a third aspect, a third is a hidden dimension, which is time. So leadership changes as the problem changes, as knowledge changes. AI changes, people change. Therefore, who leads today is definitely not going to be who leads tomorrow. So think about the example of a disease like a pandemic. In the initial outbreak, human experts led. Weeks later, months later, AI helps to identify patterns. Months later, AI automates routine monitoring. That's just a very simple example.
So leadership has to be thought of as movement. It has to be thought of as a dance. And I use the dance metaphor in the book to help get across this new way of thinking about leadership in the age of AI. At one moment, human expertise dominates. At another, AI may dominate. At another, it could be some combination of maybe mostly human, some AI, or another it could be mostly AI and some human.
Niels Brabandt EMBA MBA MSc
So would you say that there is one quadrant where you would say AI leads straightforward, all right, and human should step back? Or would you say there is no quadrant where you say this is mainly the job of AI?
Dr. Adrian Wolfberg
Well,
you know, I'm reluctant to say that there's never a role for humans, that there's a role for only AI. I think even in, you know, very routine processes, there's going to be need for, you know, routine processes maybe monitoring the status of an oil refinery, or monitoring the status of the International Space Station in terms of the amount of oxygen, monitoring critical factors in a nuclear reactor.
These are basically closed systems where we understand, for the most part, the goings into and the comings out of. And there is still going to be a need, even in those cases, for human interpretation of what that information is providing.
So I think it's, you know, I'm sure someone can come up with an example where there's no human involvement. But from my perspective in writing this book, I'm hard pressed to come to a case where there's no human involvement, zero, and 100% AI. It might be 98% AI, 2% human, but I think there will always be a human aspect required for interpretation.
Niels Brabandt EMBA MBA MSc
Excellent. So one final question here. How quickly would you say, because these quadrants are, of course, dynamic as you said, how quickly can problems move between these quadrants? Is it days, months, or even longer? How agile is it?
Dr. Adrian Wolfberg
Yeah, I mean, I think it all depends on the nature of the problem. There's no answer, prescriptive answer to that question. It's a great question, but it has to be analyzed based on the context in which you're dealing with. That's all I can say.
The leader has to be prepared for kind of scoping out what that time horizon is and may have to learn through engagement with the problem solution space what that may look like. So this is a big shift from, you know, leadership as just making a decision to leadership as being guiding the movement throughout time between these entities of human and AI intelligence.
We've not been in this space before, and we've not developed or trained ourselves to do this. So this is a challenge that we're going to have to deal with in developmental efforts for leadership and management.
Niels Brabandt EMBA MBA MSc
I think these are the perfect final words. And now, of course, the next question is, how do we navigate the human-AI collaboration? That is going to be the topic of our next episode. So for now, there's only one thing left for me to say. Adrian Wolfberg, thank you very much for your time.
Dr. Adrian Wolfberg
You're welcome, Niels.
Episode 3 of 12 (E_568): Four Ways Humans and AI Actually Work Together
Niels Brabandt EMBA MBA MSc
When AI is now in place, the question is how to navigate all of that, because now AI is there, humans are there. The question is: what happens from here? How to navigate human-AI collaboration. And we have an expert on the matter with us here today. Hello and welcome back, Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels, very much. It's a pleasure to be here.
Niels Brabandt EMBA MBA MSc
Thank you very much for taking the time yet again. And I'll pass the question straight onward to you. When someone now says, "Look, we implemented this tool, we gave it to people, and maybe they got this half-day, full-day workshop on how to prompt properly," another question is: how do we now navigate from here? How do we navigate the human-AI collaboration sustainably and successfully?
Dr. Adrian Wolfberg
Yeah, that's— that really gets to the heart of the issue. Just as a reminder to the listeners, in the last episode we asked, "Who should lead?" and we answered that. But even answering that question, there's another challenge that remains. You know, how do humans and AI actually work together, which is the focus of this podcast episode. Well, leadership is only the beginning.
So— excuse me— so human thinking, you know, begins with— before AI, excuse me. AI can never be the starting point, because, you know, people already possess knowledge before using AI. So you have to think about AI as joining an existing thinking process. So I think this is a very new concept that we don't think about. So, you know, we're already engaged in this relationship with AI through our own thinking in our own head. And the richer our understanding is, the more valuable our thinking is about, hey, how AI can solve a problem.
So it's important to recognize that AI enters into a conversation that's already taking place in our minds. So there's a value for this what I call non-AI knowledge. Well, what is this? You know, well, it's personal experience, professional expertise, contextual understanding, organization culture and history, relationships, customer understanding, political awareness, ethics and values, just to name a few.
And so what's the big deal about this? Well, because AI doesn't know what our organization has already tried. It doesn't know who supports or opposes an idea. It doesn't know the organizational politics. It doesn't know the unwritten rules. It doesn't know the stakeholder relationships and who believes in this and who believes in that. It doesn't understand what I call the contextual field.
And while it understands language, certainly the words, its leaders, managers, its humans who understand the context. And so with that, I want to introduce four human-AI collaboration modes. There are more, but I just want to, you know, deal with four because it helps us simplify a complex topic. And of course, there are spectrums between all four of these, and they can overlap and so forth.
Niels Brabandt EMBA MBA MSc
Absolutely.
Dr. Adrian Wolfberg
So one is that only you— that humans only will be the focus of solving a problem. And so this is when the characteristics of the problem are such that ethical issues dominate, the trust is critical, there might be sensitive negotiations, there might be confidential discussions, there may be unique situations.
Another— another mode is what I call AI-enhanced human. So this is probably the most common. So this is where the human asks AI, the human then interprets AI output, and that improves then the human's understanding. And a human, like a knowledge producer, would then advise a decision maker. That human remains the interpreter. That's the— the pathway of the second mode.
The third mode is what I call dual engagement. So this is a conversation between two people and two different AI conversations. This is the most interesting and most complex kind of interaction that involves three or more interacting parts. There are a number of permutations to this, but the most interesting one, the most simplest to convey, is a knowledge producer, an advisor, uses AI, but the decision maker also uses AI. Now both humans have different prompts, both humans have different outputs, and they use different interpretations. And then they communicate with each other. And then the questions emerge: whose interpretation matters? The advisor? The decision maker? Who reconciles the differences between interpretations? Who owns the understanding of the output and the interpretations that have come out from that relationship?
The fourth mode is AI only. This speaks to something we talked about in last episode. And this is sometimes very appropriate and very beneficial. It's very beneficial, very helpful for routine decisions where there's low risk and stable environments. But risks can increase. And when they do, it increases complexity, uncertainty, and consequences also increase. So AI only should be a deliberate design choice, not a default. So, you know, what changes a problem?
Niels Brabandt EMBA MBA MSc
Would you say— when I cut you off, would you say that AI only is always a warning sign then? Or is it something where you would say it could be an option?
Dr. Adrian Wolfberg
Yeah, I think that's a great observation, Niels. I would say yes, it's always a warning sign.
Niels Brabandt EMBA MBA MSc
Okay.
Dr. Adrian Wolfberg
Absolutely. I haven't thought about that exactly, but I totally agree with that. I think that's a good way to engage yourself into this shared consciousness. So, you know, one of the factors that's going on is that, as we know, AI increases the amount of information that becomes available, the speed and efficiency increases. But so do the leadership questions as well. They increase. We briefly talked about it before. Who interprets, who challenges, who integrates, who is accountable?
Well, I want to talk a little bit more about this interpretation, because I've talked about this in this episode. So AI, you know, so what does that mean? Well, AI produces outputs, just like anybody can produce outputs. But with the age of AI, it's the humans who are assigning the meaning to that outputs. Interpretation is where experience, our experience, our understanding of the context, our knowledge of things that AI doesn't know, and where judgment comes together.
Another key concept is accountability. So no matter what collaboration model we use or that exists, or what kind of combination, someone still has to own the decision. The human has to own the decision. And we have to understand the consequences and the responsibility. We have to be attuned to what might be the consequences. Well, AI never assumes what the consequences will be. Well, never assumes accountability.
So, I mean, practically, you know, before deciding on how to collaborate, you know, there's some basic questions. I provide a lot of these in the book, but what unique knowledge do your people already possess? What can AI genuinely add? And who should integrate this shared consciousness? And who ultimately owns the decision?
Niels Brabandt EMBA MBA MSc
I think these are very, very great insights here. So we see there might be a tiny warning sign when there's only AI. However, we also saw how to do it right.
The question now is, how do we then— because that's, of course, a massive change— how to adapt to change in navigating human-AI collaboration. That's exactly the topic of our next episode.
So for now, there's only one thing left for me to say. Adrian, thank you very much for your time.
Dr. Adrian Wolfberg
You're welcome, Niels. It's a pleasure.
Episode 4 of 12 (E_569): Why Changing Your Mind Is a Leadership Skill in the Age of AI
Niels Brabandt EMBA MBA MSc
AI is here, and AI is here to stay. Of course, AI means change, and when things change you always wonder, how do we handle the change?
Let's face it: no human being appears in the morning and says, "I think the best thing I can see today is change." Usually people like to keep things pretty stable, and with AI you have a massive change ahead.
The question is how to adapt to change in navigating human-AI collaboration. And we have an expert on the matter with us here today, yet again. Hello and welcome by popular demand, Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels. It's a pleasure being here.
Niels Brabandt EMBA MBA MSc
Thank you very much for taking the time again. So the question now is: we have a massive change ahead. How do we adapt to the change in navigating this human-AI collaboration? Because that is where many organizations struggle.
When you at the moment look into how many investments we have into AI and what the results are, often they are pretty modest, and the change around that is managed, let's say, mediocre at best. So how to do it better? How to adapt to change in navigating human-AI collaboration?
Dr. Adrian Wolfberg
Well, let me just first kind of recap what we've talked about in the first 3 episodes. We've talked about the leadership dilemma, we've talked about who should lead, and then we've talked about how humans and AI should, you know, work together. But all those discussion points and those answers assume the problem stays the same. And this is the real key within the age of AI, that in reality the problem almost never does.
So let's start by explaining what why adaptation is a central leadership skill in the age of AI. As AI accelerates change, the question becomes: how quickly can you, as the leader, recognize that you should change your mind? Well, what does adaptation really mean? Well, it's not flexibility. It really begins with changing one's understanding. And by changing one's understanding of reality, decisions change and actions change.
So let me begin by talking about framing, because that's the starting point for at kind of time zero when the problem is first confronted. Framing is deciding what problem you're solving, deciding what matters the most, deciding what doesn't matter, deciding where to focus attention, and deciding what success looks like. These are all human attributes. They're contextually based.
So here's an example. Just a very, very simple example. It's a business example. Sales are declining. One frame is that this is a marketing problem. And if you took that approach, then you would focus, you know, within a marketing approach, the kind of variables that would be involved.
But another frame could be this is a product problem, something about the nature of the product itself. Another frame is this is a customer experience problem. Well, these each of these frames are based on the same facts, that there's sales are declining, but they each have different solutions.
Niels Brabandt EMBA MBA MSc
Mm-hmm.
Dr. Adrian Wolfberg
So I want to talk a little bit about why frames can fail. Well, when the world changes and you have not changed your frame frame, you're going to cascade into an outcome that's undesirable.
So what causes the world what causes the world to change such that your frame has to change? Well, just using a business case again, new competitors. There could be changing customer needs. There could be new technology. There could be unexpected events. There could be regulation changes. So the organization has to learn new ways of doing things.
Well, beyond this framing aspect, there's also this reframing aspect that has to do with time. So reframing is really what I call a disciplined way of asking continually, does the way we are thinking still fit the reality we are now facing?
Niels Brabandt EMBA MBA MSc
Mm-hmm.
Dr. Adrian Wolfberg
It's a constant testing to see if change is happening. And if it is, how to adjust and reframe. And
it's a natural part of the cycle to detect what I call flawed framing. And flawed framing is a double-edged sword. On the one hand, we may associate flawed as a negative thing, but it's actually a positive thing.
Our framing will change almost all the time, sometimes a little and sometimes a lot. So realizing that flawed framing is a natural part of decision-making is beneficial. In other words, flawed framing is going to be a natural part of problem solving.
Niels Brabandt EMBA MBA MSc
That's quite a massive change for many organizations when they say that flaws usually should not be part of accepting anything in that direction. That's a pretty hard change for many leaders to say it is part of the journey, isn't it?
Dr. Adrian Wolfberg
Yes. So, you know, I think that's a critical observation, Niels. And I think it's an important message that I want to get across, that we actually want to look for and be aware of flawed framing as a positive thing. So, you know, what could be so it's just a natural part of this decision-making process in this shared consciousness environment.
So what could be the indicators of flawed framing? Well, you know, unexpected results, repeated surprises, conflicting evidence, customer complaints, employees questioning assumptions, recommendations becoming inconsistent.
So we want to make sure we as a leader, as a manager, we take a welcoming attitude, which is a totally, I think, different way of thinking than in kind of the era of the machine age, where we're focused on optimization. And flawed in a kind of machine sense is always a surprise, but in the human decision-making sense, as you say, it's not regarded in a positive way.
Niels Brabandt EMBA MBA MSc
Mm-hmm.
Dr. Adrian Wolfberg
So this kind of speaks to the question of, you know, what has previously been kind of a negative value of a leader changing their mind now becomes, you know, a positive value changing your mind. And as part of this journey of human and machine shared consciousness, solving a working on a problem.
Well, the kind of the where we start from, as we've kind of already alluded to, is that organizations generally reward certainty. So there's this notion that changing your mind after making a decision or learning something new isn't leadership. Inconsistency is frowned upon. But I think disciplined inconsistency this is what I'm talking about. Disciplined inconsistency should not be frowned upon. It's an actual sign of intelligence.
So what happens after reframing? Well, it doesn't end after we have some insight that things have changed and we need to reframe. It's only valuable if then decisions change. So we have the insight, but then we have to have the decision change. And then with considering priorities and, you know, resources, then behavior has to change as well. Action changes.
So there is a role for AI, I would assert, inside this reframing cycle. And that is that AI can help identify patterns. It can help detect anomalies and monitor change and generate alternatives and simulate options. But humans must decide, they must interpret, they must prioritize and accept responsibility.
So we can no longer talk about leadership as a static thing. We must embrace, you know, leadership as a continuous form of art, as a continuous dance. So it no longer follows this pattern of think, then decide.
Niels Brabandt EMBA MBA MSc
Yeah.
Dr. Adrian Wolfberg
Then execute, and you're finished. It is instead observe, learn, reframe, adjust, execute, observe again. It's a continuous cycle. So the practical implication for this, the practical advice is, for every, you know, important decision, we should ask, what assumptions are we still making?
Niels Brabandt EMBA MBA MSc
Mm-hmm.
Dr. Adrian Wolfberg
What has changed? What evidence no longer fits? And if we started today, would we define the problem in the same way? So this idea of continuous leadership is, again, I reinforce in the book with this idea of dance.
Niels Brabandt EMBA MBA MSc
Mm-hmm.
Dr. Adrian Wolfberg
And I think one of the values of the dance metaphor is to help expose us to new ideas for which we need some kind of anchoring, because these are new ideas and we don't know how to think about them. And so this metaphor of dance gives us some sense of this fluidity, this continuity, this engagement through time and space for being involved as a leader and manager.
Niels Brabandt EMBA MBA MSc
Excellent. And I think, of course, now the question is, what types of problems do we have and how to frame them? And that's exactly the topic which we're going to talk about in our next episode.
So you see, it is a complex issue. However, there is a solution to that, and we will show you in the next episode how to proceed from here.
So for today's episode, there's only one thing left for me to say. Adrian, thank you much for your time.
Dr. Adrian Wolfberg
You're very welcome, Niels. It's been a pleasure.
Episode 5 of 12 (E_570): Not All Problems Are Equal: A New Way to Diagnose Business Problems in the AI Era
Niels Brabandt EMBA MBA MSc
AI is here, and AI came here to stay; however, AI caused a couple of— some people say "challenges," but let's face it, there are some problems. Also, in daily business we have problems; you can label it as you like. Sooner or later, you will meet a problem.
The main question is now: what kind of problems do we have, and how to frame them? And with AI, there's a whole new perspective on this.
And we have an expert with us here today on that matter. Hello and welcome back, Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels. It's a pleasure to be here.
Niels Brabandt EMBA MBA MSc
Thank you very much for taking the time again. And of course, I'll pass the question straight on to you: what types of problems do we have, and how to frame them?
Dr. Adrian Wolfberg
Well, thanks, Niels. Let me just refresh what we've learned from the previous episodes. In the first 4 episodes we've discussed, we focused on leadership, collaboration with AI, and adaptation. But before leaders can decide who should lead or how to collaborate with AI, they need to understand— they need to answer an even more fundamental question: what kind of problem are we facing? And most leadership failures, I would argue, based on the research that I've done, begin with misidentifying the problem. And most important, understanding the problem comes before selecting the solution. That may seem obvious, but I'm just reinforcing that now.
So there's no single definition of a problem. Why is that? Well, they differ in important problems, you know, they differ in important ways. Some are routine, others have never been seen before. Some have clear cause-and-effect relationships, other have interacting factors. Some can eventually be solved, others can only be managed, not solved.
So here I want to introduce one of the key frameworks that I talk about in the book. I term it as a three-dimensional cube. Why is it three dimensions? Well, most models classify problems along one dimension, sometimes two. But reality is really richer than that. Problems have multiple characteristics simultaneously, and as they move through time. The cube idea allows leaders to see combinations rather than categories.
So I use three dimensions to begin understanding the problem. One is familiarity— in other words, what's new or not. Complexity— what's simple or complex. And wicked— where there are values and ethics are key or not so key. And, you know, within each of these three dimensions, there is a spectrum between this, between extremes. So let me just kind of go through these one by one.
So what's the difference between a familiar and a novel problem? Well, familiar problems are what we— what we've experienced, or someone else has experienced, and solved already, and we know what the solution is. Something like the annual budgeting problem, or how to process invoices, or employee onboarding, for example. So novel problems would be how we're dealing with generative AI, like ChatGPT, Claude, and so forth. Entering a new market with a new product. Dealing with a new geopolitical threat that we haven't seen before. Novel problems, because of their nature, require learning before acting. If we haven't seen something before, we have to experience and learn them before we can act. Otherwise, they're— we're going to mis— be misaligned.
Okay, so the next dimension that I talked about was between simple and complex problems. So what's an example of a simple problem? Something that has very few variables, stable relationships, predictable outcomes, predictable inputs, that we understand the process exquisitely. Complex problems, on the other hand, many interacting variables across many levels of analysis. Feedback loops. There's emergent behavior that are often unpredictable, and there's unexpected consequences.
So third— the third dimension of this cube is between wicked and not wicked problems. What are wicked problems? Well, they have no single agreed-upon definition. Stakeholders disagree about the problem. Solutions actually create new problems. And success depends on one's values and ethics as much as facts. So let me just go through some examples. Dealing with climate change. Dealing with healthcare reform. Dealing with urban homelessness. National security. And even AI governance. In wicked problems, people disagree about the problem before they disagree about the solution.
So we— I propose that we have to think about problems as hybrids, as combinations of things. So there's never a— as I start— as I began, there's never a—
Niels Brabandt EMBA MBA MSc
I have a question here. I have a question here. When— when they are hybrids, is it possible as a leader that I look at all of these categories at the same time? Because that, to me, sounds pretty complex when someone says, "Oh, it's this, this, this." And I— when I— when I probably can only reliably assess one of these categories, how do I prioritize?
Dr. Adrian Wolfberg
Well, I think you can look at all three. You have to look at each one at a time. But once you evaluate each of the three dimensions, you then have a way to think about where it is placed in this environment of— of a problem. So yes, if you were just saying, "Oh, gosh, I've got to do this all at once," that would— that is not what I'm suggesting.
You'd have to look at these one at a time, and then you could capture your understanding of the problem, and you might even name it as something unique. And the way that you can use this decision cube, you can generate almost different— different— different topologies of problems through that.
In the book, I just talk about 8, just for simplicity to get the ideas across. But there's really an infinite way. And each organization could, over time, develop its own topology or different categories of problems. And once they do that, they would then learn themselves how to do this very quickly.
Niels Brabandt EMBA MBA MSc
Yeah, very good.
Dr. Adrian Wolfberg
And what kind of solution approaches and people involved, AI involved, so they could become much more efficient over time. They would become what I called a truly a learning organization by using this. Yes, the first time you use it, you're going to be— this is new. But over time and with experience, this will become a very valuable intellectual tool.
Niels Brabandt EMBA MBA MSc
Excellent. Thank you.
Dr. Adrian Wolfberg
So when we think about, you know, hybrids as combinations, you know, this kind of goes to your point a bit, that problems are never completely one way or the other. So they're never completely familiar, they're never completely novel, et cetera, never completely simple. So every problem is this mixture that we were just talking about. And this— this mixture, this topology— these different categories that one could— one could then create is really the power of this three-dimensional cube that the organization could then use in the future in becoming a learning tool.
So let's just use an example how this may start. Let's say you want to launch a new— a new AI product. Okay, this is very topical. Well, it's— it's— the problem is familiar in the sense that you probably know your customers. Because hopefully you've created the project— product because you know the customers who want to use it. It's novel because the technology has never been used in the context that you're creating it for, since it's new. It's complex because it involves marketing, involves engineering, involves regulation, other kind of facets of an organization that have a say in what's going on. And it's wicked. The stakeholders involved with financing it or overseeing it may disagree about what are acceptable risks with this new AI product. So this is why it's important to kind of capture holistically as many of these factors that help you understand what is really going on with the problem.
Well, that's only the first— that's only the first step, because problems change over time. Well, why does that happen? And I think— I don't— I think we realize this if we think about it, but I don't think we're consciously incorporating this into our mindset. I mean, we know that new information, you know, may appear, stakeholder opinions may change, technology changes, competitors react. Because of that, our original diagnosis of the problem can become outdated to some degree or another. Sometimes very quickly, other times not so quick.
Niels Brabandt EMBA MBA MSc
And that's probably when we need to reframe problems, right? So—
Dr. Adrian Wolfberg
Exactly.
Niels Brabandt EMBA MBA MSc
And I think that's exactly our topic of the next episode. So when we see here you have problems, and sometimes these problems change over time, we have to see how do we deal with them. And the question is, why do problems sometimes have to be reframed? That's exactly the topic of our next episode. So for this one, there's only one thing left for me to say: Adrian, thank you very much for your time.
Dr. Adrian Wolfberg
Thank you, Niels. It's been a pleasure.
Episode 6 of 12 (E_571): Why Your Best Solution Today Becomes Tomorrow's Mistake
Niels Brabandt EMBA MBA MSc
We talked about the problems of AI already, and of course these problems are dynamic. They are not static; they change over time. The question is what do we do with them. And we have someone here with us, an expert on the matter, with us here today, who says problems need to be reframed.
The question, of course, I have is: why do they need to be reframed? Because for quite a while we had problems, and we had best practices, we applied best practices, and then we moved on with a solution.
So hello and welcome back yet again, expert on AI Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels. It's a pleasure to be here.
Niels Brabandt EMBA MBA MSc
Thank you very much for taking the time. And I'll put it straight to you: why do problems have to be reframed when many leaders are used to processes such as, "I have the experience," "we have problems," "we have solutions," and we just put them together and move on from there? Why do problems have to be reframed?
Dr. Adrian Wolfberg
Yeah. Well, let me just recap what we've talked about last episode. We talked about that problems have different shapes, but there's another challenge: that they don't stand still. We talked about why they evolve. That means the way we understand them has to evolve as well.
Well, one of the key aspects to help us understand this movement of change of a problem is understanding what decomposition of a problem is. So a problem, except for something that's very simple, has a lot of moving parts, typically.
And as I mentioned with diagnosing a problem through the three-dimensional cube, you can't do this simultaneously. You have to do it in parts and then integrate them, which is a much easier approach. So part of the problem understanding involves what I call decomposition in order to help us figure out the nature of the problem. Decomposition is—
Niels Brabandt EMBA MBA MSc
Can you tell us different decomposition strategies and due cause of this episode? Because how to decompose— it sounds easy. How to decompose? The question is: how do I do the decomposition when I'm not used to doing that for years?
Dr. Adrian Wolfberg
Yeah. Okay. I can go straight to that. I did want to spend a second just defining what decomposition is—
Niels Brabandt EMBA MBA MSc
Yeah, let's go. Go for it. Go for it.
Dr. Adrian Wolfberg
Yeah. Well, it's not something that's unique to AI, but it's critical in the age of AI. And that is breaking down a large problem into meaningful pieces without losing sight of the whole. You know, it's asking: how should we organize the problem so we can understand and manage it? It's not making the problem smaller; it's making it more understandable. And it helps because, you know, large problems overwhelm us.
Humans just can't process simultaneously too many variables, too many stakeholders, too many conflicting objectives, etc. So by breaking problems into parts in the right ways, we can— it allows leaders to better assign priorities, to recognize relationships, to identify dependencies, and so forth. So every sub-problem, through decomposition, has its own characteristics.
So a single organizational challenge might have a combination of, you know, routine, novel, simple, complex, wicked components. So say we want— just as an example— we want to implement AI in a hospital. Well, the technical installation could be routine. Training clinicians, on the other hand, would be novel. Patient trust would be wicked. Regulatory compliance would be complex. And the key implication of decomposition is that each part deserves a different balance between human and AI collaboration.
So there are risks in poor decomposition. Let me just give you a couple of examples. Coca-Cola: there was a time when Coke felt that it was losing the battle over Pepsi. You know, of course, that has big business implications. And the issue that Coke framed it as was a taste, a taste issue. But they lost that. They lost that because they should have included things like brand identity, customer loyalty, emotional attachment, market perception. So Coke shelved the new Coke, and they returned it to the original classic Coke. So there is one example.
The opioid crisis is another one. You know, it began with framing the problem or decomposing it into, "Well, let's just manage pain more effectively." Well, that did not turn out so well.
So there are different decomposition strategies. Before you break a problem apart, well, how you break a problem apart is going to affect the outcome. So this can be done in the following ways, and there could be combinations of doing this. You can break problems up over time through stages. You can break it up through stakeholder interests, through organizational functions, through cause and effect relationships, by geographic regions, by levels of decision-making.
And the important thing is that there's no single strategy that will work for every problem. And each of these ways of decomposing may require combinations at any particular point. And I would say there's quite often going to be the need for multiple strategies over the course of the problem solution.
So, you know, here's some practical advice. When facing a different problem, ask what smaller problems make up the larger problem. What parts that we've decomposed are routine and when? What parts are novel and when? What parts are complex and when? And what parts are wicked? And how do these all fit together?
So I just want to connect this idea of decomposition to reframing. As conditions change, the decomposition has to change as well. And that affects a need to change the frame. The collaboration is going to have to change. So between humans and AI. So every time the problem changes, one should continue to ask whether you have decomposed the problem into the right pieces.
So the book provides a framework and approach how to do this. This, you know, could sound overwhelming at first, but the book takes you through, kind of step by step, how to think through this. And once an organization does this a couple of times, they will be in a much better position to do this more efficiently and more effectively over time. And stay ahead and not become defaulted to AI output.
Niels Brabandt EMBA MBA MSc
Excellent. And of course, we now should wonder, hmm, why is then human thinking still needed? And that's exactly the topic we're going to have in our next episode. So is human thinking still needed? In which cases, and in what form? We're going to talk about that in a minute. So for this episode, there's only one thing left for me to say: Adrian, thank you very much for your time.
Dr. Adrian Wolfberg
You're welcome, Niels. It's been a pleasure being here.
Episode 7 of 12 (E_572): AI Changes What Kind of Thinking Matters, It Does Not End It
Niels Brabandt EMBA MBA MSc
AI is here, and AI is here to stay. And many people wonder, well, do we still need humans? Is human thinking still needed? And—well—we will find an answer quite soon, because we have an expert on the matter with us here today. Hello and welcome back, yet again, Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels. It's a pleasure to be here.
Niels Brabandt EMBA MBA MSc
Thank you very much for taking the time again. And I go straight into it: why is human thinking—is it needed? And if yes, why is it needed, and if not, why not? I know it's a really tough question, and I'll pass it on straight to you because it's a question that moves many people in today's times.
Dr. Adrian Wolfberg
Yeah, no, I think it's a really fundamental, fundamental question. Let me just recap where we've been that led up to this question. We've spent the last 6 episodes understanding problems, how to frame them, but that naturally leads to this question of this episode: if AI is becoming so capable, well, what is the unique role that's left for human thinking? Well, I want to address that, you know, directly. I want to explain why human thinking is always needed.
So there are a number of ways that humans contribute, and these are not—this—I'm giving you a list here of a few, but, you know, it's a partial but important list. We know purpose about things. We have an understanding of values. We have an ethical perspective. We understand context, which is critically important. We have experience. We have organizational memory. We have empathy, which is critically important for understanding consequences. And we own accountability.
So let me set—let me set the stage for what AI does that we don't do well. Well, it has exceptional capability for speed, also to scale up. It's exceptionally good at pattern recognition and information synthesis. So those—so I've kind of laid out some very powerful ways of why human thinking is needed.
So, let's see, to help undersuring—to help ensure that we understand the problem, in the book I talk about a framework that allows us to look at it from kind of 3 different parts, 3 different kind of dimensions. So how—how we characterize a problem involves asking 3 questions. What kind of problem is it, which we've talked about before, but also the next one is, how is the problem structured? And then the third is, what kind of human thinking is required? We've talked about the first 2 in ways. I want to focus on this—this third way, this third dimension.
So this is where both critical and creative thinking come into play. So critical thinking is one of those terms that we—we banter around, but we rarely define it, so it's kind of nebulous. But what it involves is the ability to evaluate evidence, to question assumptions, compare alternatives, test conclusions, and identify weaknesses in an argument.
It's also where creative thinking comes into play. Creative thinking, same. It's used ubiquitously but rarely ever defined. But—so this—the characteristics of creative thinking are: ability to generate possibilities, reframe problems, connect unrelated ideas, imagine alternatives, and seeing opportunities.
Well, I want to talk a little bit about structure. Why does a problem's structure affect our thinking? Well, structured problems—what are they? Well, there's known rules about how to engage that problem. It's got clear objectives, repeatable processes. They need more evaluation and less inventions, less imaginative thinking. Unstructured problems, they're uncertain, they're evolving, ambiguous, and there's a need for more imagination and more exploration.
So I just focus on the structure because it leads us to justify and point to the kinds of human thinking that are needed in a problem, and which are this creative and critical thinking. And they often work together, critical and creative thinking. Every important problem requires both, some less and some more, and at different times. You know, creative thinking creates possibilities when that's needed. Critical thinking tests them when that's needed. Critical thinking asks, "What if?" Excuse me, creative thinking asks, "What if?" and critical thinking asks, "Will it work?"
But there's often a reassignment over time. So a problem may begin requiring creativity. Later it requires evaluation, critical thinking. And so later it becomes routine, minimal human thinking, which has an implication, obviously, to the whole human-AI collaboration combination.
So, you know, there's—there's a common misconception: many people and organizations assume that AI replaces thinking. But really, instead, I strongly argue that AI changes which thinking becomes more valuable. Routine analysis, routine thinking declines in value. Judgment, creativity, interpretation become more valuable.
And as I mentioned earlier, as AI becomes, you know, better at generating information, routine thinking becomes less valuable. But the abilities to frame problems, to imagine alternatives, to exercise judgment, and integrate competing perspectives become more valuable than ever. In other words, AI doesn't eliminate human thinking at all. It changes the character of—
Niels Brabandt EMBA MBA MSc
Good news, yeah.
Dr. Adrian Wolfberg
—of, yeah, of which kinds of human thinking create the greatest value. It's—it means that AI varies what I call the economics of thinking.
Niels Brabandt EMBA MBA MSc
Excellent. So I think this is pretty good news here, because I think many people fear, "Oh my God, is human thinking going away?" And it sounds like we definitely will need it for the foreseeable future, no matter where AI is going. We will always need it. Would you agree with that?
Dr. Adrian Wolfberg
Well, yeah, because these human attributes of meaning-making, interpretation,
understanding the context—all these kind of human attributes—I don't think are ever going to be replaced by AI. You know, the logical extension of that argument would be when humans become robots, and that would be the end of, you know, human civilization.
Niels Brabandt EMBA MBA MSc
Yeah, yeah, that's something we definitely do not want to happen. Absolutely.
Dr. Adrian Wolfberg
It's kind of—
Niels Brabandt EMBA MBA MSc
So very good.
Dr. Adrian Wolfberg
Right, right. It's kind of an absurd, you know, and, you know, to think about that way, I don't think we'd ever go that way, to be completely robotic. But
the—yeah, so anyway, it's—
Niels Brabandt EMBA MBA MSc
Yeah, we can agree human thinking stays, and I think these are the perfect final words here. Of course, we now have to see when human thinking stays, and yes, it is needed.
We still need to have something we just call adaptive capacity. The question is how—what is the competency that we need to build that, and that's exactly what we're going to talk about in our next episode.
So for this episode, there's only one thing left for me to say: Adrian, thank you very much for your time.
Dr. Adrian Wolfberg
You're welcome. It's been a pleasure, Niels.
Episode 8 of 12 (E_573): The Leadership Skill Nobody Trained You For: Adaptive Capacity
Niels Brabandt EMBA MBA MSc
AI is here, and it came here to stay, but of course we now need something which is the competency for building adaptive capacity. The question is: what is that, and why do we need it? And we have an expert with us here today. Hello and welcome back, Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels. It's a pleasure to be here.
Niels Brabandt EMBA MBA MSc
Thank you very much. So what is the competency for building adaptive capacity?
Dr. Adrian Wolfberg
Yeah, before I get into that, let me just recap what we talked about in the last episode. We explored the kinds of human thinking that AI cannot replace, but thinking alone isn't enough. For the next 3 episodes, I'm going to address the leadership competencies needed for framing and reframing. These, I argue, are new competencies that we're not yet— we've not yet developed, and that are needed on the horizon to begin developing.
Well, first— well, the first one is: leaders have to learn about changing conditions, and that's where adaptive capacity comes in. So what do I mean by adaptive capacity? This is not about, you know, being flexible. It's about the ability to absorb new information, recognize when old assumptions no longer fit, to update our understanding, and to change direction without becoming disoriented.
So I want to also say that this reframing that we're— is our, you know, objective to be able to do, to become adaptive. We need to learn. We need to be able to learn first. So you can't reframe until you've learned something new. So learning— what does that mean? It means there might be new evidence, new perspectives, new relationships, new explanations. Without learning, our original frame remains frozen.
So there are two diagnostic questions that can help us move towards this reframing, toward this adaptive capacity. How overwhelmed am I by information? Modern leaders rarely suffer from the lack of information.
Niels Brabandt EMBA MBA MSc
Mm-hm. Rather too much of it. Way too much of it. From all— here, something here, you have to consider this, and all of it at the same time. Could you please decide by tomorrow? Yeah.
Dr. Adrian Wolfberg
Yeah. This is what we call information overload. You know, competing reports, competing deadlines, endless dashboards, constant notifications. The question is: can I actually process all this? To what degree can I?
And the second diagnostic question is: how understandable is the information that I'm receiving? Sometimes there's plenty of information, but there might be multiple competing interpretations. There might be conflicting expert opinions, contradictory AI outputs, different stakeholder perspectives, uncertain evidence. And this is when understanding becomes harder than just collecting the information itself.
So when we look at these perspectives, these dimensions of the amount of information and the level of understanding, we can actually see kind of four categories. And again, I just use the four categories for simplification of getting across this idea. It's not meant to be a statement that there's only four, and there's only one at a time, and so forth.
But, you know, as an example, if there's high information overload and high clarity, this could be something that is typical of routine financial reporting. If there's high information overload and low clarity, this could— we could see this in the early stages of a global pandemic.
Niels Brabandt EMBA MBA MSc
Mm-hm.
Dr. Adrian Wolfberg
Where there's low information overload and high clarity, this could be some simple operational decision. When there's low information overload and low clarity, you might see this in an emerging geopolitical crisis. So I want to just, you know, introduce these learning modes. In the book, I provide a framework that helps leaders recognize what learning challenge they are facing. Not every situation requires the same learning strategy. Some require more discussions, experimentation, filtering, etc.
What happens is that AI changes learning. What AI dramatically increases is access to information, summaries, pattern detection, alternative viewpoints. But it cannot determine which interpretation is most meaningful. AI helps you gather information faster, but it doesn't automatically help us learn faster.
The danger of information overload, which we all kind of intuitively and experientially understand, is that we simplify too early. We ignore conflicting evidence. It's too painful to do that. We rely on familiar assumptions. We stop questioning. I think we've all seen this. We've all experienced this. So it's ironic that more information can actually reduce learning. That's the message— one of the messages here.
And the same thing with multiple interpretations because of the lack of clarity. You know, these interpretations can arise because of different experiences that we have, different incentives, different contexts, different values. So leadership isn't choosing the loudest interpretation, which I'm sure we've all seen that. It's really about evaluating competing interpretations.
So if information is overwhelming, AI can help organize that. But if interpretations conflict, humans become the increasingly important factor in that. So in kind of practical advice, when learning stalls, ask yourself: am I overwhelmed? Am I confused? The lack of clarity. Do I need more information or a better interpretation? Is AI helping me understand, or is it simply giving me more to read?
Niels Brabandt EMBA MBA MSc
Excellent. We see here: humans are still needed. And you see here: it's okay to feel overwhelmed and confused as long as you get the right conclusions out of that and then move on from there.
The question now is, of course, as it gets more complex, there's an important skill which you call mental acuity. A very, very great term, and that's exactly what we need here.
However, that's the topic for our next episode. So for this episode, there's only one thing left for me to say: Adrian, thank you very much for your time.
Dr. Adrian Wolfberg
You're welcome, Niels. It's been a pleasure.
Episode 9 of 12 (E_574): Mental Acuity: The Leadership Skill That Turns Information Into Understanding
Niels Brabandt EMBA MBA MSc
When you have AI in your organization, what kind of leader do you need? You probably now say someone who is probably charismatic, or smart, or knowledgeable on the matter, or who's able to deal with all of it at the same time. To lead us forward somewhere toward a sensible and reasonable innovation.
Well, we have a better term for that now: mental acuity. However, there's a lot more to that, and we have an expert on the matter with us here today.
Hello and welcome back, Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels. It's a pleasure to be here once again.
Niels Brabandt EMBA MBA MSc
Thank you very much for taking the time again. I'll put it straight to you: what is the competency of mental acuity? What's behind that, and what does it mean?
Dr. Adrian Wolfberg
Well, I'm— yeah, I'm going to answer that. Let me just recap the last episode. We talked about adaptive learning, which was, you know, how we absorb and update our understanding. But this kind of learning alone is not enough for a leader. Once information enters our minds, we have to process it, and that's where what I call mental acuity comes in.
So many people hear the term mental acuity and they— they may automatically think this is about your level of intelligence or your IQ, but I mean it in somewhat of a different term. It's the ability to notice what others overlook. It's recognizing meaningful patterns that are important to you. It's being able to distinguish signal from noise. It's reflecting before reacting. It's recognizing when assumptions need questioning. It isn't about being smarter. That's, I think, really an important point here. It's about thinking more clearly.
So what is— so there's this direct link between mental acuity and meaning-making. So as we talked before, information by itself has no meaning. Whether it comes from AI or whatever, humans have to— we, whether it comes from a human or AI— we have to interpret what it means.
And so I want to spend a few seconds to focus in on what meaning-making is, because it's an important characteristic that only humans have. So it comes from understanding context, from understanding personal and collective experience, from understanding the operative relationships and the status of those relationships, from being able to compare various factors, whether it's human or not. And this idea of interpretation, which we talked about in previous episodes. So it's just a very simple example I used before, this declining sales graph. The graph doesn't— it tells you what's happening at a macro level, but it doesn't tell you why.
Niels Brabandt EMBA MBA MSc
True.
Dr. Adrian Wolfberg
People have to create the meaning for that information. So I next want to talk about uncovering hidden patterns. Well, we've— I've mentioned before that AI is great about identifying statistical patterns. Well, what kind of patterns do humans identify? We recognize organizational patterns, behavioral patterns, political patterns, historical analogies, and ethical implications.
So AI detects patterns in data. They're great at that. But humans detect patterns in meaning. And I think this is an important facet of this episode to reinforce this idea of human pattern detection, what that actually means. It's a detection of meaning, of patterns in meaning.
I also want to talk about reflection. Reflection deserves very special attention, one, because it's most often neglected. And it's most often neglected because of time constraints. Either we attribute those constraints internally to ourselves, or externally through other people, through the organization, through systems outside the organization.
So what— let me just— what is— what is reflection? Well, it's slowing down. And what is— so what? Why do we have to slow down? Well, in order to do interpretation, we have to examine assumptions, whether they're internal assumptions to ourselves or assumptions outside of ourselves. We have to have the opportunity to connect experiences. We have to ask better questions, ask bigger questions. And we have to reconsider first impressions.
These are, I think, all very justifiable reasons why reflection is important and the implication of having reflection. The first reaction might be, "Oh, well, this is going to slow things down." Well, I would argue that it's not going to slow progress. It's moving direction so that progress has a better chance of achieving the desired outcome.
So we can think about mental acuity across two dimensions. The first is attention, which we've kind of talked about, internal and external.
Niels Brabandt EMBA MBA MSc
Mm-hmm. Yeah.
Dr. Adrian Wolfberg
So what is this internal attention? Examining our own assumptions, monitoring our biases, evaluating our reasoning, recognizing emotional influences. Then there's the external: observing reality, gathering evidence, noticing environmental changes, listening to others. Leadership requires both internal and external attention.
And the second dimension is cognitive functioning, which spans the spectrum from processing to clarifying. So what's processing? Receiving, organizing, and absorbing new information. Clarifying, what's that? Well, making sense of integrating information, testing, interpreting what has already been received. Processing gathers information. Clarifying creates understanding.
So there are these four modes that are most commonly observed in terms of mental acuity. The book goes into these in detail. Again, I— for simplicity, I just talk about four. There's more. They operate in conjunction with each other over time. But the four modes emphasize what we observe: our reflection, how we explore, and integrating existing information.
Each becomes valuable under different conditions. Sometimes conditions warrant multiple of these. So the point is, you know, not to remain in one mode, but to know when to shift.
So how does mental acuity affect reframing? Well, poor mental acuity misses weak signals. It overlooks assumptions. It accepts first explanations and repeats familiar thinking. High mental acuity detects change earlier, questions existing frames, integrates competing evidence, and supports better reframing.
So when we talk about mental acuity in AI, you know, AI does the great, you know, function of collecting, summarizing, comparing, and generating. Humans, the interpretation and so forth. So as AI increases the amount of information, mental acuity determines whether more information actually produces better understanding.
So from a practical advice perspective, many people assume with more information, better decisions. Well, I argue that what's really going on, what's more— what's a greater success factor, is better processing in the mind, internally, externally, results in better understanding. With better understanding, then comes better decision.
Niels Brabandt EMBA MBA MSc
I think these are the perfect final words here, especially the summary at the end, how to spot great, high, and low mental acuity. I think this is extremely helpful for organizations listening here.
Of course, now you need to be able to organize probably across different knowledge topics at the same time. And then we need a skill which is called boundary crossing. And we're going to talk about that in our next episode.
However, for this episode, there's only one thing left for me to say: Adrian, thank you very much for your time.
Dr. Adrian Wolfberg
Thank you very much, Niels, for having me. Appreciate it.
Episode 10 of 12 (E_575): Why AI Might Be Making Your Silos Worse
Niels Brabandt EMBA MBA MSc
AI is here, and AI is here to stay. We talked about mental acuity; the question now is: how do you actually deal with, let's say, different knowledge areas and topics? And we need a skill here which is called boundary crossing. The question is: what is that? What is this competency? And we have an expert on the matter with us here today. Hello and welcome back, Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels, for having me. It's a pleasure.
Niels Brabandt EMBA MBA MSc
Thank you very much for being here again. So, what is the competency for boundary crossing? Because many people know, especially from the corporate world, how they keep their boundaries and say, "This is my job, and that is exactly my job, and I stay within this boundary." So why suddenly do we talk about boundary crossing? So what is the competency of boundary crossing?
Dr. Adrian Wolfberg
Yeah, that's a great lead-in, Niels. First, let me just do a recap of the last couple of episodes. We talked about how individuals learn, how individuals think, but organizations solve problems through groups of people. They rarely do it through one person. So that introduces a new challenge: how do we connect across different kinds of expertise to a common understanding?
This is where boundary crossing comes into play. So what do I mean by knowledge boundaries? I mean those things that we as humans create distinctions, and we create them between things. And Niels, you kind of already alluded to the general idea.
So, for example, we create a distinction between engineering and marketing. They do very different things. They process very differently. Clinicians and administrators, knowledge producers and policymakers, lawyers and technologists, finance and operations. So these are just examples of different people who have different kinds of knowledge. And each group, more than what I've just said here, uses different language. They value different evidence, and they define what's important for success so they define success differently.
And so it's terribly important as to why this thing of boundaries is important. So organizations, from my research and what other people have done, is they rarely fail because people know too little. Yes, that can happen, but they're more likely to fail because experts don't understand one another. Information remains isolated in silos. Assumptions aren't shared because they're protected because of resource constraints. And perspectives never or rarely get integrated for similar reasons for protection and resources.
So the key is making these differences surface to make them understandable and usable. So boundary crossing isn't asking everyone to agree. It's helping different experts, not different people with different knowledge, understand, respect, and use each other's knowledge. So, for example, an engineer sees technical feasibility. A marketeer sees customer perception. A lawyer sees regulatory risk.
Niels Brabandt EMBA MBA MSc
Compliance, yes.
Dr. Adrian Wolfberg
None are wrong. But a leader, somebody, a leader has to integrate these. So we can look at this construct of knowledge boundaries through a couple of dimensions. The first is the spectrum between process and people. So what do we mean by that? This is a process. I'm going to talk about the two ends of the spectrum, and obviously things can occur in between.
The process end is where we're emphasizing systems, workflows, consistency, repeatability, and efficiency. That's an emphasis on an extreme end of the spectrum. The other end of the spectrum, the extreme end, is people focus. So what does that mean? Well, we know what that means, but you know, relationships, communication, trust, motivation, collaboration, morale. And organizations need both. Processes move people, move the work. People move understanding. So that's just a dimension. I've laid out the two ends of the spectrum.
The other is a spectrum between implementation and understanding. So at one extreme end, implementation asks, how do we execute? How do we deliver? How do we make this happen? So it's all about how questions. The other end of the spectrum is understanding. What are we solving? Are we solving the right problem? What does this really mean? What assumptions are we making? These are all what questions. So those are ends of the spectrum between implementation and understanding. And organizations need both, but implementation without understanding leads to efficient execution of the wrong problem.
So you can see this two-dimensional framework, and you can already imagine, you've probably already experienced, where the organization has emphasized too much on one side or the other and not included the right amount of the other end of the spectrum. So in my usual style, I have four modes. Again, I create these frameworks for simplicity by identifying four modes. There's obviously more. I think these are important. They can be experienced in combinations, but I just lay them out individually. So different situations require different emphasis.
So I have these kind of four modes that you need to think about. One is coordination. One is facilitation. One is translation. What translation, what I mean by that, is to be able to interpret the language that's used by someone in one organization so that the person in another organization can understand what they mean. You know, we actually have unique words that we associate with our own processes. So it's a translation and a transformation of the language. The other is integration of the different kinds of knowledge from the different boundaries.
So this framework helps leaders recognize when a bridging capability is most needed. So if you think about this in terms of its impact for AI, while AI can connect information across disciplines, it's only leaders that can connect people across disciplines. So that's an important nuance. And there's perhaps a counterintuitive result. AI, I argue, can increase organizational silos. So it can actually increase the problems that hierarchies have.
So what do I mean by that? So engineering asks AI one set of questions. Marketing asks another set from their perspective. Finance asks another. Legal asks another. Each receives excellent answers from its own perspective. Without boundary crossing, these perspectives are going to drift apart. They're going to stay separate. Counterintuitively, that means that AI can strengthen these silos unless leaders deliberately reconnect the people in these silos.
So this boundary crossing, this boundary competency is going to become even more important. I mean, it's important now. It's been important in the machine, in the industrial revolution age, which we're now shifting, and even in the information age. It's going to become more important in the age of AI. So how this, so practical advice, before making important decisions, you know, ask who hasn't been part of the conversation? What expertise are we missing? Have we translated our assumptions so people in other disciplines and other organizations understand them? Does everyone define the problem the same way? A key question.
So how this connects to reframing. Reframing improves when our perspectives are integrated. So poor boundary crossing creates partial frames, competing frames, and contested frames.
Niels Brabandt EMBA MBA MSc
I think these are the perfect final words here. You see how important boundary crossing is. And probably when we have people from corporations sitting here, you might remember these moments where you all contributed very well, and at the end, they were all single islands drifting apart, and you wonder how could that have happened that six months later, nothing has happened. We now see how to do better.
The question is now how to put the leaders' design into practice. And that is exactly what we're going to talk about in our next episode. But for this episode, there's only one thing left for me to say. Adrian, thank you very much for your time.
Dr. Adrian Wolfberg
Thank you, Niels. It's been a pleasure.
Episode 11 of 12 (E_576): From Theory to Practice: Designing an Organisation for Human-AI Collaboration
Niels Brabandt EMBA MBA MSc
You've heard quite something from boundary crossing and lots of other topics so far. The question is: how do you now put it all together into the leader's design? The question is: how to put the leader's design into practice. And we have an expert on the matter with us today. Hello and welcome back for episode 11 out of our series of 12. Hello and welcome back, Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels. It's a pleasure.
Niels Brabandt EMBA MBA MSc
Thank you very much for taking the time again. And of course, I have to ask you, well, how do we now—so first, of course, what do you understand under leader's design, and how do we put the leader's design into real-world practice of our today's organizations?
Dr. Adrian Wolfberg
Great. It's a great topic to talk about, Niels. Let me just recap what we've talked about. Throughout this series, we've explored the leader's dilemma, how to frame problems, how humans and AI collaborate, how leaders adapt in the age of AI, and how competencies are needed to make that work in the age of AI. Next comes, as Niels said, the practical question: how do we embed this in an organization?
So now I want to shift a bit and talk about organizational design. But I don't mean about creating organizational charts. What I'm talking about here is much deeper. It's thinking about design processes, communication patterns, governance, incentives, routines, meaning structures, and learning systems. And here is where change management plays a critical role. Many AI initiatives fail. I would say many—any initiatives fail—because organizations focus on deploying technology rather than changing behavior. So it's the behavior that is critical here.
Successful adoption of AI is going to require new habits, new expectations, new decision routines. And most important for our episodes here is new leadership competencies and behaviors. Technology adaptation of AI without organizational and behavioral adaptation rarely succeeds.
We've already seen three in the research. We've already seen three organizational errors with respect to AI. These are warning signs. Very early on in our series, we talked about the immediate default to AI as a warning sign. I just want to talk a little bit more about three major categories of warning signs, which this topic of immediately defaulting to AI comes into play.
And that's—the first is framing errors. So this is when the organization solves the wrong problem because they misunderstood the problem. So, for example, their focus is on optimizing efficiency when trust between people is the issue, or when automating a symptom instead of addressing the cause. This would be going back to Laura's observation that AI may immediately be the default solution when we have not addressed the cause. This would be an example of that.
The second category of error is what I call alignment errors. So this is when everyone agrees on the problem, but they may not agree on its priority as a problem. They may not agree on the objectives in solving it, or the assumptions behind the problem, or the responsibilities of people in the organization to participate in solving the problems. So different parts of the organization drift apart. We also talked about that in your example in the last episode, I think.
And the third is, I think—I don't know if it's the most important—I think it's the most visible and can have the most consequential impact on us, and that's suppression errors. So suppression occurs when dissent disappears, when we're uncomfortable providing evidence. Because if we do, that evidence is ignored. When questioning becomes discouraged, there's risk associated with questioning. When all recommendations are accepted, which may sound counterintuitive, but without challenge, meaning that they're not really going to be thought about and action taken upon.
So this suppression is very complex. It's really critical for leaders to be on top of and to contribute to resolving because it comes from organizational culture. It comes from the nature of hierarchy. It comes from being overconfident, which can be supported by the organizational culture. And it can come from over-reliance on AI, which kind of speaks to the case of an immediate default to AI. So those three errors, I think, are important to have in our back pocket for leaders.
So let me just talk about, from a change management perspective, this idea of institutionalizing these aspects. So framing and reframing need to be institutionalized, this kind of behavior. So they have to make this behavior, framing and reframing, an expected habit and not just an occasional activity. So how do you do that? Well, leaders, managers have to model this behavior. They have to act with this behavior in mind so others can see them.
So, for example, at the beginning of a project, the leader can state explicitly how they're being framed, the problem is being framed. They can explicitly talk about the assumptions while revisiting the problem during reviews. They can encourage alternative perspectives. They can ask whether the problem has changed. They can conduct a kind of after-action review focused on framing and reframing. And so what is I'm arguing that this type of behavior is repetition. And I'm saying this type of behavior is a strength.
So when we continuously model behavior, we're repeating this model. That's what that means. And it's not a negative thing. It's a positive thing. So what happens is that many leaders search for the perfect framework. But leadership is built through repeated questioning. The frameworks are only there as a guide. The goal is to have the right questions repeatedly as reality change. So the repeatedly is the repeated modeling of behavior.
So this idea of what is the role of leadership is it's a role that I would call a journey. There's no finishing line. Because organizations, they continually face new tech, new markets, new regulations, new competitors, new crises. So leadership is a continuous adaptation, which is not just the sole responsibility of the leader. They have to model this behavior throughout the organization. And that, of course, is a leadership challenge.
And again, remember that framing is a uniquely human responsibility. We create the frame. AI can analyze it, predict, recommend, optimize based on the frame. But we determine the purpose, the priorities, the acceptable risk, the ethical boundaries, and so forth. Organizational life is not technical. It involves trust, politics, culture, relationships, competing interests, emotions, and value. So these cannot be optimized using algorithms. We're part of a social system. We're not part of a technical system.
So practical advice: every leadership team should ask, have we framed the problem clearly? What assumptions are we making? Has anything changed since we last discussed this? Is everyone working from the same understanding? What evidence are we overlooking? Who feels safe enough to challenge our thinking? So leadership in the age of AI is no longer about having the best answer. It's about designing the conditions under which the best understanding can emerge.
Niels Brabandt EMBA MBA MSc
I think these are the perfect final words here. So you now know how to put the leader's design into practice. The question now is, what do leaders need to do now? Because we now talked about a lot of topics, and the question is how to bring it all together.
So the big final episode, number 12 out of 12, is coming up in a minute. First, we're going to have, of course, this episode to reflect on it, and then we move on from there.
So for this episode, there's only one thing left for me to say. Adrian, thank you very much for your time.
Dr. Adrian Wolfberg
Thank you, Niels. It was a pleasure.
Episode 12 of 12 (E_577): AI Does Not Change Strategy. It Changes Leadership. Here Is What to Do Now.
Niels Brabandt EMBA MBA MSc
AI is here, and AI came here to stay. In 11 episodes we talked about different aspects of what you need to do in the age of AI: as a leader, as an organization, as people involved in leadership, as people working with leaders and amongst leaders.
The question is: what do leaders need to implement right now? And we have an expert on the matter with us here today for the big final episode, 12 out of 12. Thank you very much for coming yet again, Dr. Adrian Wolfberg.
Dr. Adrian Wolfberg
Thank you, Niels. It's been a pleasure.
Niels Brabandt EMBA MBA MSc
Thank you very much for taking the time for us here again. So, after everything we discussed here, what is it that leaders need to implement now?
Dr. Adrian Wolfberg
Well, let me just—before I get into that—recap the journey that we've been on.
Niels Brabandt EMBA MBA MSc
Yes.
Dr. Adrian Wolfberg
For those who may not have captured all the episodes. So, over the last—this is the 12th episode—over the last 11th episode, we've explored one central theme: AI doesn't simply change how organizations use technology; it changes how leaders need to think. So we began with the leadership dilemma, we examined who should lead, we explored how humans and AI collaborate, we learned why framing and reframing are critically important, we explored the competencies leaders need, and we finished by examining organizational design to implement changes.
So, for humans, AI is not the transformation—yeah, okay, there's technological transformation, we've, you know, talked about that—but for humans, the transformation is leadership. So organizations, unfortunately, ask the first question: how do we implement AI? But the better first question is: how should leadership change because AI exists? Why? Because technology implementation is temporary, but leadership redesign is long-lasting.
So what do I mean by redesigning decision-making? Well, redesigning decisions: we do this by explicitly framing problems, deciding who leads, defining when AI contributes, creating pause points, revisiting assumptions, assigning accountability. We have to keep judgment at the center. Yes, AI contributes information, analysis, and recommendations, but humans contribute judgment, context, values, ethics, empathy, and responsibility. That means that judgment becomes infinitely more important as AI becomes more capable.
So we have to make reframing a leadership habit, leadership habit. But to make it a habit, to ask: what is changing? Does our original frame still fit? What assumptions should we revisit? Don't wait until failure forces reframing. Make reframing routine. Build adaptation into the organization. Well, adaptation comes from learning routines, reflection, questioning assumptions, integrating perspectives, encouraging experimentation. It's not simply hiring adaptable people; it's a design competency.
So another thing you can do is invest in leadership competencies. We need to develop leadership competencies that strengthen human capabilities in the age of AI. This is one of the biggest messages here. We need to look at adaptive learning, mental acuity, boundary crossing, critical thinking, and creative thinking. Three or more of the episodes have been focused on this. These become strategic assets to the organization. So you can see that human thinking becomes immensely and more critically important in the age of AI, not less.
Another thing that you can do is protect human thinking, which is an outcrop from what I just mentioned. There's going to be this tendency to default to outsourcing AI to human things. We want to avoid that. We want to avoid outsourcing AI to framing. We don't want to outsource AI to judgment. We don't want to outsource it to interpretation or for ethical and value reasoning, and for empathy, empathetic reasoning. We want to use AI to strengthen thinking, not to replace it.
Another thing is to encourage a culture of questioning. I mean, this has always been a message that management scholars have talked about, but in the age of AI, this has become infant—infantescently more important. So you need to normalize, through modeling behavior, questions like: are we solving the right problem? Are we—what assumptions are we making? What evidence challenges us? What changed since last week? So questions should become organizational habits.
Another behavior is: treat human-AI collaboration as dynamic. It's not just a one-of or a one-case-of. Because who leads changes. Problems change. Tech changes. Organizations change. And much more. Therefore, leadership must continually adapt. And so I want to end with looking forward. The future belongs to organizations that learn continuously, that adapt confidently, that integrate machine and human intelligence, and that maintain human responsibility.
So, in closing, most conversations about AI have focused on what the technology can do. You see this in what's written about, in blogs, in social media, and in books and articles. Throughout this series, I've focused on something different: how leaders should design the conditions under which people and AI think differently. Technologies will continue to evolve, but the need to frame problems, to exercise judgment, to integrate perspectives, to adapt to changing reality will remain. And those enduring—those are enduring leadership responsibilities. So thank you for joining me throughout this podcast series. I hope these conversations encourage you not only to use AI differently, but to lead differently.
Niels Brabandt EMBA MBA MSc
I think these are the perfect final words. There's hardly anything for me to act—uh, to add here, because you see exactly what Adrian, Dr. Wolfberg, just said.
We often see that people say, "What can technology do here? What function do we have here? What prompt can we do there?" And it's a lot more than that. Leadership was, and always will be, the key to sustainable organizational success.
So at the end of this whole podcast series, there's only one thing that's left for me to say: Dr. Wolfberg, Adrian, thank you very much for your time.
Dr. Adrian Wolfberg
Thank you, Niels, for hosting me on this great, important topic. I appreciate it.