Daniel Serfaty: Welcome to MINDWORKS. This is your host, Daniel Serfaty. We’re kicking off season three today with a focus on leadership from the science of how we understand leadership to leadership in practice and in business. We are also going to focus on exploring how the COVID pandemic has shaped and changed the landscape for leaders and for leadership teams. So today I have the honor of talking to two guests that are deeply versed in the science and evolution of what we define today as leadership.
First Dr. Stephen Zaccaro is professor pf psychology at George Mason University, where he is a director of the industrial organizational psychology program. He’s also an experienced leadership development consultant, and an associate editor for the Journal of Business and Psychology. He also serves on the editorial board of the Leadership Quarterly, the Journal of Leadership and Organizational Studies and Personality Science. He’s a fellow of the Association for Psychological Science and of the American Psychological Association, especially with divisions that are relevant today, division 14, so society for industrial organizational psychology, and division 19 about military psychology.
Also joining me is Dr. Ronald Piccolo. He’s department chair and Galloway Professor of Management in the College of Business at the University of Central Florida. He’s currently serving as a special assistant to the University of Central Florida president Alexander Cartwright, leading the development of UCF’s five year strategic plan. From 2009 to 2016 he served at Corner Professor of Management and academy director of the center for leadership development and executive education in the Crummer Graduate School of Business at Rollins College in Orlando. Ron has also been a visiting scholar at Drexel University in Philadelphia, Tulane University, Jacobs University in Bremen Germany and the Amsterdam Business School.
So welcome Steve and Ron to the MINDWORKS podcast. I’m absolutely delighted. I have followed your publications and your major contribution to the science of leadership.
Ronald Piccolo: It’s great to be here, thank you.
Stephen Zaccaro: Thank you, Daniel. Glad to be here.
Daniel Serfaty: So please introduce yourself to the audience, I probably did not do justice to your extraordinary accomplishment, but I am more curious to say, what made you choose this particular field of endeavor? You probably had a choice of any field in psychology or management science, and yet you focused a lot of your scholarly research on leadership. Maybe we’ll start with you Steve.
Stephen Zaccaro: Thank you. I’ve been doing studying leadership for over 40 years. I guess I got started with it graduate school. I was at a social psychology program at the University of Connecticut when social psychology was decidedly not about groups, but I was interested in that topic. So I began to reach some more in it, and I began to do some dissertation work on groups and took a class on groups, I did a paper on leadership and that really started it. I’ve always been fascinated by social dynamics and people working well together, and always believe that leaders and leadership is perhaps a primary driver of that effectiveness, of that collaboration.
So that’s where I ended up. My first paper in the area was on leadership and continues to now. So it really started with an interest of mine in graduate school that was different from what the whole field was doing at that time, and just kept pursuing it. It’s also how I ended up at organizational psychology because we were studying organizational psychology, we were studying leadership more than in other parts of psychology.
Daniel Serfaty: We all benefited from that insight that you had in graduate school Stephen as I had the privilege of working with some of your former students and I saw how much they learn with you.
Stephen Zaccaro: Thank you.
Daniel Serfaty: Ron what made you choose this field?
Ronald Piccolo: I was involved and had the opportunity for several leadership activities as I was growing up as a kid, even in college, I was resident advisor and residence hall director, I was on student government so I was active as a emerging young man in leadership. But I was a math major. I wasn’t in the business school, I wasn’t in psychology, I was completely a mathematician. But it was in graduate school also, I got a master’s degree in business, then when I started the doctoral program, even though I started the program on the macro side of the management field, trying to study strategy, there were several faculty members in our department who were focused on organizational behavior and psychology, and group work, and leadership.
They just were wonderful people, wonderful scholars and my inherent interest in leadership was sparked by them. And once I got to work with them on their research and see how they did it and what the science of leadership meant, and how it encompasses so many different fields, how so many different academic traditions are revealed in the leadership process, that became immediately exciting to me and then kind of continued to go from there.
Daniel Serfaty: There you go. For the audience, you can start with mathematics or with social psychology and end up in the leadership business.
Ronald Piccolo: Well I joke if you don’t mind, I don’t mind making the joke. One of the things I learned as a mathematician is there’s an average, which means there’s above average and below average. And pretty sure I was below average in math. That was pretty clear. So I think the path for me out of that particular field was made clear as I approached the end of my undergraduate degree. I’m like, yeah I could do some math, but there’s some people in the room are all fully lot smarter than I am, I might want to try something different.
Daniel Serfaty: Did you find that mathematical or rigorous quantitative training helped you later?
Ronald Piccolo: I joke about it, kind of pure mathematics and that’s its own thing, but I have surely benefited from a mathematical way of seeing the world and trying to develop, we’re still developing, but trying to see patterns, see contingencies, see substitutions. That’s one thing that I remember so vividly in trying to develop a mathematical proof. Like the answer to the problem is not right in front of you, it’s not necessarily what’s right there, but you have to maneuver, you have to manipulate, you have to substitute, you have to look at things in a slightly different way and then the solution to the problem can emerge.
It’s not calculus, differential equations or linear algebra purely, but the process of thinking through problems and solutions and substitutions and alternatives and detail and contingency and all that, that really allows people to be good scientists and good thinkers, and good problem solvers. So I make fun of myself a little bit about mathematics, but make no mistake, I benefited greatly from all of that training.
Daniel Serfaty: Steve you came at it from a social psychology perspective and understanding how human interactions contributes to the group, or the organization, to the team. There is almost here in, especially in America, some fascination with great leaders, especially in business. But in general we all had a leader in mind or a model of a leader, and it can range from Lee Iacocca, from Chrysler or Napoleon and anything in between. Was there during your studies, one particular leader that created an aha moment for you or an insight, particularly? We’re going to dig into the theories of leadership a little later, but I want to tell for you, whether or not there was an example that you were … mental model almost that you-
Stephen Zaccaro: You’re asking me about a example of a great leader and I’ve thought of so many over my career. I’ll just talk about the one I’ve been thinking about in the last couple years, and that goes back to the Renaissance in Italy and Lorenzo de’ Medici and just fascinated how he created a climate for fostering the Renaissance, and the integration of idea and the explosion of new ways of thinking. He evidenced what we now call a transformation leadership style.
He also evidenced other kinds of styles of leadership, which is a key point of I’m thinking about leadership. So, I’ve been fascinated in how he operated as a leader in Florence and the great changes that he did, it’s a fascinating story. And how he used to gather scholars around the table who would disagree with one another and he was comfortable with that disagreement. Innovation comes out of that if it’s done right. So if you’d asked me earlier in my career, I might have said some other things, but lately I’ve been fascinated by his leadership style.
And you brought up why are we fascinated by leaders? It’s because when a leader gets it right, it has major impact. I’ve looked back at leadership, I went back to antiquity and many of our stories, our memes, our stories, our legends are about leadership. And we keep repeating those stories down to the centuries and through history because of the great impact. I think that’s why we’re fascinated by leadership is that, when someone gets it right then we see great impact. That’s what’s always fascinating me about leadership. They can have such tremendous influence, but they could also be very destructive, obviously. So they have great, great influence on what happens.
Daniel Serfaty: Whether it’s on the positive side or on the negative side of it so, yes. In a sec we are going to dig a little deeper into that, but first I wanted to explore two topics, and then plant the seed and come back to them later. Because that’s one of the topics during our podcast series that came almost as a, if we talk about artificial intelligence or robots, or human work, or the future of work, that thing comes back again and again, everything is changing in a world that is fully connected.
We are not at the Lorenzo de’ Medici’s environment in a world when we have networks everywhere, when the access to information is media, that is inherently decentralized with automation, intelligent device, is the role of leadership changing in a sense? Are they perhaps less important than they used to be? We’re all fascinated by leaders. Who wants to talk on that question?
Ronald Piccolo: I’ll start Steve, if you don’t. I think your question is about the role of technology development, whether it’s automation or AI, or the remote world that we find ourselves in here. And I know Steve’s got some ideas on this. I think my opinion, or at least what is heightened to me, is what’s always been true about leadership, which is the pattern of conundrums that leaders face. Which is to say, okay on the one hand we want leaders and we can look at leaders for their effectiveness and their ability to move organizations, that’s on the one hand. On the other hand, leaders are dependent on what other people do, right?
And so here’s a conundrum. We are going to hold you accountable, but if you don’t actually necessarily do what’s theirs, that’s one example of a conundrum. Another conundrum which I think is exaggerated by technology a little bit, which is to say, how customized should we be? Or how personalized, or how individualized should we be with the members of our group, or the people we’re trying to influence such that I am willing to adjust and adapt for every particular need?
And how do I balance that against creating a consistency adhering to tradition, stability, et cetera? And so again I think the world we find ourselves in now, exaggerates the tension that exists between those two pursuits. On what things do I have to apply the same rule for everyone? Non-negotiable, we all do it the same way and we come in, and on what part of that do I let people use their own judgment and empower them? At which point do I hold control and authority on particular issues? And what point do I let other people make it?
And this is not new. This is not suddenly a concern, suddenly a conundrum but I do think the world of work and technology exaggerates that. Because a lot of people can make their own way, people are reevaluating what work means, they’re reevaluating what their role is, reevaluating what their identities are when it comes to work. So your question is about, how has leadership changed? I’m not sure yet and I have to give more thought about this. I’m not sure that this changed yet, but it certainly has exaggerated some of the natural conundrums about leadership. That’s my thought on it at the moment.
Daniel Serfaty: That’s very interesting. Emphasize certain aspects of leadership more than others as a result of this. Mostly I would say the hyper connectivity we find ourselves into. Steve you want to add to that?
Stephen Zaccaro: Yeah I do. But I thought Ron those are excellent points. I want to take it in a different direction in terms of, Daniel you said leadership becoming less important with increases in technology. I said, no, I think it’s going to maintain its high importance or become more important. What does technology do? And broad sense it increases information flow and therefore, from different sources, and therefore it increases informational complexity. A lot more is coming into the team, into the organization, into the top leadership.
And it also, as we see in COVID, increases what I would call social complexity. You’re connected with many more stakeholders. You are relating to people in different ways through different technological modes. And leadership is often about creating some kind of organization and stability within the context of chaos, within the context of complexity. So, looking back in the last two years, when everything hit and we went virtual and had to learn all of these new systems, I noticed how much more people who were very competent, very skilled, relied on the leaders around them to help get through it and help them provide direction, and for motivation. Motivation became very important as people became discouraged by the contact.
So, when you increase complexity through technology, which is what we’ve done for centuries, it doesn’t replace leadership, it just makes leadership that much more important to help guide the collective, the organization, the team, through that greater complexity. So that becomes more natural for them. And we always seem to say leadership is now less important. Now, when I say leadership, I don’t mean just a leader, I think we’re going to talk about this later but, leadership is shared, leadership can be a network. Could be multiple leaders working together.
I just want to throw that out there because you started talking about the hero again when I mentioned Medici. Leadership is much more complex and in some ways technology, Zoom, and these interconnectivity devices that we had, the multiple ways of connecting has made more leadership as a phenomenon, as a forum, more complex. So yeah, I think that’s what technology has done. Increased the need for leadership, the importance of it, and also has changed the kind of forms it could take.
Daniel Serfaty: I totally share both of your perspectives here, even as a corporate leader myself, I’ve seen the past several years, not reduce the importance of leadership, and I agree with you, we need to talk about shared leadership because most people associate leadership with that charismatic person on top of some organization, which is a bit distorted, but it has emphasized certain aspect of what we do. And reduced perhaps some other aspect in a sense, we started to play a different role, especially during the pandemic over the past two years.
And we’ll go back to that notion about many of our audience, many leaders in our audience are wondering the degree to which they have, or they should modulate their behavior as leaders as a result of the extraordinary stress that has been infused within the workforce. So, thank you. We’re going to spend a little time now, go back a little bit to the beginning and trying for our audience, we have two senior academic leaders here with us and I want to go back to really the roots of leadership. What is it, et cetera. And then we’ll expand about the future of the leadership.
But I’ve read recently an extraordinary number, I don’t know if it’s true, these days you read stuff and you have a way to validate. That there are 15,000 current books in print about leadership, or related to leadership. I guess they may not have leadership in the title, but they’re related to the leadership. It’s pretty extraordinary if that’s true, but even if half of it is true, it’s still extraordinary. Why is that? Why is a topic so popular?
Stephen Zaccaro: I think it goes back to what I said earlier, if I could jump in, because leaders can have great impact positive and negative. I saw recently a study that said that there’s the biggest stressor, or the biggest empowerment in the workplace is the supervisor. Tremendous influence on one’s life. And when we look at the course of events and we look at course of history, we find that individuals or leadership processes have enormous impact in terms of what happened. So I think that’s why it’s so popular because of the tremendous influence leaders can have on followers at multiple levels. They can do great things, they can do terrible things.
And they can excite and motivate large numbers of people, or they can demoralize and really hurt the same numbers of people. So, I think the impact that leaders and leadership can have is what drives those books, what drives the interest. It’s one of the reasons, as I said earlier, I’m in the field.
Daniel Serfaty: It’s an impact thing. Ron do you agree?
Ronald Piccolo: Yeah I do agree and I also think having done some scan of those books, not all 15,000, a lot of the ideas are similar in those books, but they’re written for slightly different audiences. So, there’s leader in politics, in business, entrepreneurs, celebrities, athletics, or sports, you name it. There’s different takes on it and I agree with Steve, when we tell stories about why companies or groups, or organizations or teams succeed or don’t, a natural explanation for that is the activity and the influence of a person who’s been charged with overseeing activity in a certain direction.
Daniel Serfaty: Well, let’s dig about that. So let’s review basic definitions of when you teach leadership to your graduate students or at the university, what is the basic definition? What is leadership from your perspective? And I think we need to differentiate the leader from the quality of leadership. What is leadership? I’m challenging you probably with something you’ve done for combined 70 years here perhaps, but for our audience, how would you define what leadership is?
Ronald Piccolo: Steve I’ll let you go first on that.
Stephen Zaccaro: I was thinking of Ralph Stogdill, who’s a famous researcher who said, “As many definitions of leadership as people studying leadership.” In my own career I think I’ve wrestled with three or four different definitions. Ron and I probably agree that at its core, it’s social influence. Whenever I teach about leadership at a basic level, or I’m giving workshops about leadership, I will start off by saying, the basic fundamentals or the functions of leadership is to provide direction, and manage others around that particular direction.
I think about it as goal directed social influence that instigates a collective action. So it really comes down to either providing a direction, or facilitating a direction emerging from a collective so that people are … that there’s agreement on that. And then orchestrating how that direction or that goal is going to be accomplished. That’s a very functional view of leadership but it just incorporates so many other things that we talk about. When people ask me I say, if you want a place to start, start with those two functions and we can go from there.
Ronald Piccolo: I’ll be specific because I’m in the midst of working on a book right now, and two books, one of them has a leadership chapter and this idea comes through. So I completely agree with Steve that it’s a process of social influence, probably because we share a similar mindset. But of course that takes leadership at the level of human interaction, it’s not leadership of a enterprise, or leadership of an organization, that’s a little different. But the definition that I’m using relies on Joseph Rost, I don’t know if you remember him, Steven, he had a book some years ago, Leadership for the Twenty-first Century.
But he defines leadership as an influence relationship among leaders and followers who seek to realize a shared purpose. And again, embedded in that definition are a number of important things where there is reference to leaders and followers, not just the leader, which highlights the kind of interdependence that exists between these two entities. I think it’s foolish to define leadership only about what the leader does, the purpose of the engagement, and the people who are being led.
Steve may have taught me this where you ask the class, what’s the most important thing a leader must have? And you get all kinds of answers. Of course the answer is, they have to have followers or else they’re not leaders, they’re just standing out front, wandering around. So the definition that includes an influence relationship among leaders and followers that seek to realize a shared purpose includes the importance of vision, the fact that it’s shared, meaning that the followers have bought in. They’re highlighted as a key active participant in the influence relationship.
So they could have as much influence on the leader as the other way around. So for me personally, that is about as close to the best definition, or the definition that’s most consistent with my own view.
Stephen Zaccaro: We’re both emphasizing the goal directed or purpose, but implicit in there, and this kind of thing operates at all levels, at the top of organizations, at top of nations, all the way down to a project team. You’re providing direction and you’re orchestrating or managing that, but the relational aspects are critical. Leadership happens in a system, followers and leaders and networks, and it’s not static the relationship. I think this is one of the things that recently leadership researchers have been really been thinking a lot about is that the dynamics of the leadership relationship can change.
So that maybe you are formally a leader and there are formal followers, in a reporting relationship, but informally over the course of a project, over the course of an episode of things happening so followers, so-called followers can engage in leadership activities. Different leaders can step forward at different points. So, leadership is much more complex, it’s certainly more complex than just the individual leader. You have a leader and follower, but even that relationship could change. There’s a number of really smart leadership researchers who are diving into this phenomenon over the last decade or so.
Daniel Serfaty: That’s fascinating because it paints a portrait much more complex definition. Some early teacher about our leaders are born or they are made, you probably are working with your students on that. And what would you have to say to our audience this time to look at the person of the leader? And it can be a situational leader. It can be a leader that is a leader for a few days, or passing that on. Can we make leaders? Can we develop leaders? Or there is some traits, I may be saying a word that is not very welcome in the literature these days, are there some traits that define a leader?
Stephen Zaccaro: That’s where my work on leadership started. My first paper was titled An estimate of leadership variance due to traits. And we were using that word at a time when, no, it wasn’t welcome. I will hasten to add that traits are certainly part of the leadership conversation since then. We certainly talk about traits. We talk about stable dispositions of the person that contribute to leadership. When someone asks me, it’s a common question I get, “Are leaders born or made?” And I say both. You were born with a temperament and there’s a lot of research that supports the genetic temperamental basis for leadership.
You’re born with an orientation toward leadership, whether it’s in a motivational value, somewhat of an orientation. Your experiences, your development, starting in childhood, which is some of the things I’m investigating now, shape that more, move it more toward a direction of leadership and how to … Ron, you talked about your early experiences as a leader. I mean that shaped your orient. I believe you were oriented that in the beginning. So, leaders born and made is both.
There are people who do not have a temperament, they don’t want to be a leader. They may have some of the skills, but they don’t have that fundamental motivation to drive to be a leader. So it’s always a more complex question. And traits matter, we know that now, but it’s much more complex than that. And I think it’s hard to pin it down to traditional traits. That was my first aha moment when I was doing research, we actually were looking at the situational approach. We looked at studies that manipulated the situation, changing the membership of the team, changing what the team had to do. And the assumption was that, when you did that, a different person would emerge as leader for that situation.
And to my surprise, going way back to what the major way of thinking was at that time, the same person emerged across these different situations. How do you explain that? Because traits mean you behave in consistent ways across different situation. And what I zeroed in on, early on, and I keep coming back to that idea 40 years later, is that the attributes or traits that are important for leadership are those that foster that ability to change as the situation requires. The ability to recognize when change is necessary to be agile and flexible in how you respond. I think that’s what the great leaders do.
But that’s not a traditional trait that says, you’re an extrovert so you’re going to be extroverted across different situations as I know how I can change my response to different situations. So for me that created a much more complex view of traits and situations and how they work together, and how leadership happens.
Daniel Serfaty: How wonderfully subtle those distinctions are. I like that very much. The notion of adaptability being embedded already in what the leader does, and therefore being able to move from situation to situation and still be a leader is very-
Stephen Zaccaro: And I will point out that many of the constructs we’ve been studying over the last couple of decades now, emotional intelligence, political savvy, organizational savvy, adaptability, social intelligence, they all have that at their core. The idea that you look at a situation you recognize what may be needed and you are able to adjust your response to that. So to me, that’s what ought to be driving the discussion and traits and so forth.
Daniel Serfaty: Ron, do you want to add to that again about-
Ronald Piccolo: I mean it’s brilliant. He’s right where I think we need to be as a field, it’s a little unsatisfying in some ways, for some people, particularly when they want a prescription, like how should I lead and what should I be doing? Or what should my style be? So it’s unsatisfying to suggest, well you need to be adaptable and flexible, and attentive to the demands of the situation. But that’s really the reality of it. In terms of whether leaders are born or made, I’ll turn it back to you. Are musicians born or made? Are artists born or made? There’s some artists who have a natural inclination and a vision and a way of portraying their reality, that’s fantastic and wonderful.
And others who learn, learn the proper techniques all along the way and are effective and functional and develop beautiful art. So if leadership is an art, then there are people who have a natural instinct for it, but certainly you can learn all the things that are worthwhile. And if you’re with me on stories here, I want to pick up on something that Steve said, an aha moment for me, I know that was something that we said we’d get to at one point was an interview I watched with Mike Krzyzewski, who’s the coach of the Duke Blue Devils College basketball team, you may know him.
I want to say this interview had to be maybe 10 or 15 years ago. He’s being interviewed about his season and maybe they just won the championship and they’re talking about how he goes about practice. You can tell the reporter is trying to get behind the curtain. Like what is he doing? Why is he so successful? And one of the questions is, “How are you going to coach this year’s team? What’s your plans for coaching this year’s team?” And he said, “Well, I don’t know yet.” And the reporter is like, “What do you mean you don’t know yet?” He’s like, “Well, I don’t know what kind of team I have. I don’t know how gritty they are, I don’t know how cohesive they are. We’ve just recruited some high profile freshmen, I’m not sure how well they’ll play, I don’t know how they’ll adapt to our system. And so not sure how I’m going to lead yet until I see what I’ve got and then I’ll know how I have to lead.”
And I was like, wow, that’s brilliant. Because here’s one of the … we hold him up like a hero as one of the great leaders and he’s admitting straight away, I don’t know exactly what I’m going to do, I just have to adapt to how this situation presents itself and I’ll make my choice from that point forward. And that was an aha moment for me, and it highlights, I think, the interaction and the interdependence that exists between a leader’s instinct or natural preference in style and what the team, or those being led allow to happen, if that makes sense.
Daniel Serfaty: That’s a wonderful story about this ability to adapt in this case, to the, not just to the particular skills of basketball, technical skills of your follower, but also to the team interaction and the kind of team he’s going to have there. Quite a big take on that theme. So Mike Krzyzewski in this particular case is a basketball leader, coach and adapted from year to year, depending on the kind of … how he’s going to make lemonade, depending on the lemons he’s going to have. But the question is that, is there evidence in your research, or in the literature of those leadership skills, or competencies, or traits that transcend the domain?
Can a great military leader becoming great leaders in industries, for example? And where does it work and where does it fail? Do you have some example of the transfer, or skill transfer from one domain to another, whether it works or not?
Ronald Piccolo: Steve what’s your thought on that?
Stephen Zaccaro: I think the fundamental skills of leadership do transfer from situation to situation, from domain to domain. Because leadership fundamentally is about problem solving, social influence, leading and motivating others, and we could talk about a range of fundamental leadership activities that we see in different domains. Now that doesn’t mean that the same person can be a very effective business leader and then you pop them into coaching a team for example, or running an academic. I mean there’s been many instances of failures for people, for example with great military leaders, then you put them in charge of a business and it fails.
And to me that is more about not understanding the changes in the context. Number one, how the context is changing, two, also basically the knowledge. Leadership is a cognitive exercise. A lot of leadership is about thinking and understanding your context, and that takes the basic knowledge of that context. So when I hear about, can you be a, the question, can you be a great leader in one setting and other settings as well? If you don’t have the core, basic knowledge of that setting, then that’s where you may stumble unless you build a group of experts around and who can do that.
So, on one hand, I just believe that the skills and capacities of leadership are the same across different domains, but the actual solving of particular problems within particular context needs more than that. And that’s where you often see leaders who try to go to different situations stumble a bit.
Daniel Serfaty: Yes. And sometimes you see that the stumbling, so to speak, even within business, but people who shift from one type of industry to another. As you said, there is that knowledge component, but there also the whole internal dynamics of an industry may be very different than another, and to go back to adaptivity, adaptability, that quality that the leaders should have to, not only to understand that they’re in a different situation, but also to having them what it takes to adapt to that new situation.
Stephen Zaccaro: Yeah, exactly. And I think that’s where I’ve seen stumbles, where a leader comes in, from outside, comes into an academic setting, for example, and assumes that the context is the same. I’ve watched a lot of military leaders go into the civilian sector and say, “Wait, this isn’t like the military.” The failure there is not to recognize its shifts in the context of the situation, understand what that means for your leadership and adapt accordingly. Now there’s so much in that statement. I said understand what that means for your leadership.
I will just add that a lot of leaders don’t think about their leadership. When I work with executives and leaders, I often finish my first session with them by saying, my homework for you is to be curious about your leadership. Think about your leadership, understand who you are as a leader. Because that’s how you can then understand how you will interact with situations and how you’re thinking about situations. And a lot of leaders aren’t self reflective. The great ones are. When I’ve met some leaders or they talk about it, they can expound on what leadership is, and can be humble about their own leadership, as well as recognize where they could be great. And that helps them get into situation.
When you crash into a situation and say, this is like my old situation, I was great in my old situation and therefore I’m going to apply it here because I was great in that situation and it crashes and you blame them, when maybe you needed to be a little more self reflective and understand how your leadership display needs to change as a function of that situation. Lot of qualities, leadership qualities in there that we need to be thinking about.
Daniel Serfaty: Absolutely. I am smiling inside because I’ve seen a couple of those transfer, even on my own personal experience. But I remember working with the military with a four star Admiral who was literally the commander of the Pacific forces. So he had half of the earth under his command, literally. And he was a great military leader. He literally commanded all the forces in the Pacific ocean. And he retired at an early age, like many of those flag officers retire relatively early, early enough to enter industry. And he was miserable in industry because he grew up, he spent 30 years in an environment when things happen, when he gives a short order and things just happen. And that’s not the reality in industry.
Even though he understood that, he was smart enough to understand the difference, he didn’t have in him what it took to adapt to the new reality. So that’s a good example. Ron do you have some example of successful transfer or unsuccessful ones?
Ronald Piccolo: Well we see what you just described from military to industry, we see it from industry to academia. So people who’ve been in industry and come to the academic setting and they are frustrated by the pace, by the number of people they have to get involved and how it all works. I have been in several conversations with donors, or stakeholders, or people who want to have something to say about the university and they’ll blurt out, “Why can’t the president just tell them what to do?” I’m like, yeah, yeah, I see it doesn’t quite work that way.
As you pointed out early in my introduction, I’m working on the strategic plan for the university. And we had a moment where I’m sitting with the board of trustees and faculty have come to listen and share their feedback on the process so far, and I don’t necessarily have to tell you how faculty can be in those settings. And the trustee leaned over to me and said, “Well that’s why I don’t let employees get involved in my strategic planning in my company.” And I was like, if the faculty were employees, the way your people are employees, then yeah, that would be true, but it’s not the same.
And to lead that way and have expectations, and demands, and the nature of the circumstance, it’s just not the same entity. And so I think to your point about the military transfer, and I think to Steve’s point about recognizing how, the willingness and the ability to recognize how the situation and the demands, and the networks, and the politics and the pockets of power are different from one situation to another, the willingness to see that, the ability to recognize it, and the humility to adapt to it is so critical. Because, I think it was Steve who points out, people will rely on how great they were yesterday as explanation for why you should listen to them today.
And it’s like, yeah, that’s great that you did that over there, but this is a different issue. To get back to the original question, I have always admired Steve’s early work, Steve, on the functional role of leadership, like leadership as a function, and leadership’s role in group functioning. It’s nice to talk at a high conceptual level about the philosophy of influence and so forth, but most of us live with the idea that leadership has a purpose. When someone’s in that role, or people assume the leadership role, it is for a particular function.
We have to coordinate activity, we have to create certainty where there’s not certainty. Or we have to navigate the complexity, as Steve pointed out, of information, or of accelerated competition or whatnot. So the leadership role has a function. And it’s not necessarily the responsibility of one person, there could be a dozen people in the group who can fulfill the function of leadership to provide clarity, to help with coordination, to overcome the agency problems that always exists in groups.
And so I think when people have transitioned from one business to another, or one situation to another and have been successful, had the ability to recognize how the situation has shifted, and the humility and the willingness to a shift accordingly, and to have a mindset of like, what is the leadership for? what am I expected to be doing here? What is needed of the leader now in this situation at this time? And am I willing to do that? Because again, we can all point to examples of people who, you know, think we’re supposed to do this when everybody else wants to do that. And that’s a disaster.
Stephen Zaccaro: Yeah. I wanted to jump in on that, first of all I love that expression you gave Ron, humility to adapt. The concept of humility, the trait of humility is really a focus of a lot of recent leadership research and humility adds a powerful leadership trait, a quality. And I think it’s because of this, because it may contribute to adaptation, the humility to adapt. Ron just talked about my focus on leadership functions and so forth, I want to go back to what he said earlier about relationships.
If you notice in the mishaps and destructions that we talked about of leaders moving from one sector domain to another, what is changing is the fundamental nature of the relationships. So the nature of relationships in the military are different than in the private sector, or even in the civilian sector of the government. And the nature of relationships as Ron and I both know, is very different in academic settings from business and so forth.
So when people stumble, as we said, it’s a failure to recognize how the nature of the relationships have to change. And this goes back to what Ron mentioned earlier that, at its core leadership can’t happen without followership and social influence can’t happen without relationships between the influencer and the influencee, if I may add. And just to make it more complex Ron mentioned, groups of people can lead. And that means you must understand what the relationships are among those people, and how those relationships can shift so that one person steps up to take the leadership role informally, even within a formal system.
To me, that’s what’s critical about where leadership is going. So humility to adapt and shifts in the nature and type of relationships across situations is which I think adds to the complexity of doing leadership.
Daniel Serfaty: I like very much the conversation of the past 10 minutes because it shows how in a very tangible, practical way, insight from your research could help leaders talk about this notion of influencing and relationship management. And I love that phrase too, the humility to adapt, and empathy and all these other key aspect or key dimensions. We’ll be back in just a moment, stick around. Hello MINDWORKS listeners, this is Daniel Serfaty. Do you love MINDWORKS but don’t have time to listen to an entire episode?
Then we have a solution for you, MINDWORKS minis. Curated segments from the MINDWORKS podcast condensed to under 15 minutes each and designed to work with your busy schedule. You’ll find the minis along with full length episodes under MINDWORKS on Apple, Spotify, Best Proud or wherever you get your podcast.
So I want to devote the last part of this portion of our podcast to question of, how do we develop? What does academic research do or can help us? For people in the audience they are thinking, okay, I’m the head of an organization or of a team, how can I develop leaders? Or how can I develop leadership skills in those people that are in my organization? And perhaps an associated question that both of you brought up earlier, you talk quite a bit about followership.
And how do we develop also better followers? So it’s not what they are but how we take them from where they are to make them better leaders, and also eventually better follower. What can academic research contribute to our understanding of leaders development.
Ronald Piccolo: I’ll take a quick stab at this Steve, on the leadership development piece if you don’t mind. My experience in trying to teach leadership in the classroom and with executives and with students, if you will, two things seem to have the most influence on when people develop, and develop their skills. And one is just experience, people having experience or not, or in the classroom, or in executive coaching, it’s either trying to create experiences for people so they can see how they react in different situations, or get their licks a little bit, like repetitions if you will.
That’s like being an NFL quarterback in a way, it’s like we can watch your video of what it means to be a quarterback, or you could take a snap and see how you do. So I think the experience is valuable, and in some cases for students, we have to put them in situations and create experiences for them. And Steve’s done executive coaching over the years. Another option is to help people interpret the experiences they’ve already had. So kind of like allowing people to think back over the moments when things have gone well or not, and break those down, deconstruct the circumstances.
What was the situation? Where were you? How did you prepare? What did you think of? What were you worried about? A little bit of deep clinical psychology to like, okay, I’m not going to necessarily be able to put you in a situation, but you’ve been in situations so let’s figure that out. So from my point of view, how do you develop people is creating experiences or allowing them to have a deep interpretation of experiences they’ve had. And the other is figuring a way, and maybe it’s, again I’ll go back to the expression of humility, but allowing people to develop a heightened sense of self-awareness.
There are a lot of tools that we use, right? There’s psychological and personality assessments and individual differences, and there’s plenty of highly valid and reliable indicators of tendencies, preferences, competency skills, and they’re valuable, particularly if you can interpret them properly, and how to make sense of it. But we use those with all our students and sometimes it’s just like being able to put a label on what people are doing. It’s like see, this is what we mean. Your tendency to have a high level of ego strength, for example, your default interpretation of the situation of like, what does it mean for you?
Remember when you told me that story and your immediate reaction was, I was worried about what this meant for how I looked? Okay, that’s an indication of your ego strength. And now that we’ve labeled that and we know what that means, let’s explain how that could be interpreted, how that can affect what you do and how you influence. So long answer, I think my experience tells me, create, or allow people to reflect on experiences and then try to figure out ways to elevate and heighten people’s sense of self awareness. Particularly for their instincts and their natural tendencies. Because then that allows them to grow. That sort of captures and summarizes where they are at the moment, and then we can build on from there.
Daniel Serfaty: That’s a very wise and practical advice actually, I like it a lot. Steve you want to add to that development.
Ronald Piccolo: Yeah lots of good stuff there. So I got to build on that. I think we have to think about leadership development as a system with multiple parts to it. And Ron alluded to that so, yes we have to give them experience or experience based learning, but a colleague of our Scott [inaudible 00:47:28] and his research folks have published papers on the fact that experience will work, but only if it’s coupled with feedback. And you have to have feedback availability. So where does feedback come from? So it comes from the boss, others learning partners.
Ronald Piccolo: So that means you need to develop experiences, but you also need learning partners who are going to help you prompt the insight into what you’ve learned from it. Ron mentioned assessments and self-reflection, you have to … experience without reflection, what have you learned from it? Is almost useless. So it’s not only feedback but it’s that assessment part. So you see how these parts work together. And the reflection is critical, absolutely critical. That’s why I say, be curious about your leadership, development is about changing behaviors, and adding to your skill repertoire, and understanding how they apply to different situations and maturing your complexity as a leader.
Ronald Piccolo: We can’t change personality. We can’t take an introvert and make them an extrovert. But we can take introvert and say, how could you act as an extrovert in some situations? And what does that mean for you afterwards? We can also take someone who is introvert or someone who’s highly conscientious, for example, and say, how does that orientation, that personality of yours, affect how you do leadership? So someone who’s very conscientious gets into the details and so forth, gets an assessment from a 360 that says they’re micromanagers and they won’t delegate.
Ronald Piccolo: And so you might say to assessment, hey, this orientation which has lots of good strengths, good stop, look how it may be hurting you and how can you manage that? So development, that self-awareness is so important because again, it’s about how you are as a leader, how you act as a leader, how you can change how you respond to different situations. So assessment self-reflection, self-development, learning partners, experience, even the role of formal courses and workshops, so that’s another part of the system.
Ronald Piccolo: I think Ron you mentioned this earlier so I want to elaborate it. One of the things we do in development is we teach behaviors and people can come out of that saying, “This is the stick I have, the hammer I have, that I will beat every leadership now with.” And that’s not good. The point of the workshop is to add that skill to your repertoire, but what is needed is a further understanding of when you might be able to use it when not. So I get worried about a lot of workshops and training programs that tend to be behavior specific, and give the message that this is the hammer you could now use. This is what you should be doing, which is the opposite of effective leadership as far as I’m concerned.
Ronald Piccolo: That’s where experience comes in and learning partners, and reflection, and so forth. So to me development is that system of multiple strategies that need to work together. And I will say, most people who do leader development don’t really do this well. We may run a formal program without bringing experience so we may bring coaches in. If you subscribe to leadership development as a system in an organization, and you’re head of HR, or you’re head of this, you have to think about all those components.
Ronald Piccolo: How do you foster the right experiences for the right people? How do you foster the feedback and learning partnerships that can get the most out of those experiences? How do you foster self-reflection from those experiences? I don’t know that companies and organizations spend enough time thinking about development as a system and how they can foster all parts of it.
Daniel Serfaty: I love it. It has a sound of truth because all I can think about is my own experience. And both of you just gave me a couple of pointers about how to add to that system, how to add components to that system. Personally I’ve been very suspicious of, if there are 15,000 books, there are probably 20,000 different seminars about how to develop leaders and the five steps, the five secret steps to become a great leader. The three secret components of leadership. And your remark over the last 10 minutes about leader development shows a much more subtle, much more individualized, much more experiential feedback oriented ways to shape leadership, which is what real life is. And so thank you for sharing that with our audience.
Ronald Piccolo: I’ll do a sequence to your point. In class I have a slide deck, it probably has about 40, maybe 45 book covers of leadership, and I kind of go through all of them, kind of setting up what you described. It’s like here are the five secrets to management. And these authors know five things that really make a difference, but they don’t know the seven things because these guys know two more things than those five guys do. But neither of them know as much as Jack Welsh, who’s the 29 secret so be careful here because you might be selling yourself short if you only get the five secrets because somebody has a few more.
Daniel Serfaty: Thank you for revealing your own secrets today to our audience, they make imminent sense. You just share with our audience really very insightful and thoughtful ideas about how to think about leader development at a system leader development as an individualized approach, but typically we all alternate, as you mentioned earlier, between the roles of leaders and followers depending on the situation. It takes skill to be a good follower. And the question is that are we, as academia focused on that and thought about to define what are those skills and how to best develop them? Even for people who alternate roles be between leaders and followers depending on the situation.
Ronald Piccolo: So the question is about the concept of followership and the role of followers in the process. I think Steve and I agree that followers are and should be treated as active in the influence process, not as passive and obedient. They never really are and so we shouldn’t think of them that way. I think if we go back to the definition that leadership is an influence relationship among leaders and followers, that tells the followers that they have an active role in the process. And that they have an influence on how the leader sees the situation, understands the situation and reacts.
Ronald Piccolo: They are also engaging with other members of the same team and that has an influence on how things unfold. And so, part of the notion of followership is recognition of the influence that the followers have on the dynamic of the relationship, and on the group’s outcome. So there is some literature, probably about 15 years or so, about followership as a concept, and I think it’s sort of an evolution of the relationship approach to leadership, a heightened tension to the influence relationship among leaders and followers and that has spawned examinations of the follower’s role.
Ronald Piccolo: And so this is a critical thing and the literature kind of points out the importance of self-awareness among the followers, their own identities, how they see themselves. Identity is a big issue with a lot of these things, like how do you see yourself? How do you see your role? What is your responsibility here? You draw the example of a military commander, in that setting that person has a particular identity, and a particular view of what his responsibility is, his expectations are, and that reflects on how he behaves.
Ronald Piccolo: But moving into another role, they have a different identity of what they’re expected to do and how they should react. Some of the research that I’m personally familiar with is understanding the follower identity. Do the followers think they’re expected to be obedient, passive and just sort of take it? Or do they see themselves as active contributing participants in the relationship? And of course any good relationship you got to sort out who’s expected to do what. That’s what I am most familiar with about the research on leadership and followership, so I’m not sure where this conversation ends other than to say, Steve and I, if I could speak for you Steve, we agree about the importance of active followers in the relationship.
Ronald Piccolo: And what the literature is starting to point to is appreciating, recognizing, and helping followers appreciate and embrace their roles as such.
Stephen Zaccaro: Yeah let me jump in a little bit. Let me start by talking about briefly the history of followership work. Back in the ’60s I think Ed [Hollander 00:56:17] talked about the critical role of followers in recognizing a leader, granting legitimacy to the leader. Just because you’re appointed to a formal leadership role doesn’t mean you can effectively practice leadership, you need the followers who will say, “Yes, I will follow you with motive, with energy. With energy; I will follow you with energy because I see you as an appropriate leader.”
That’s what Hollander introduced that idea and Bob Lord, Robert Lord really expanded on this with leader prototypes and followers having their heads and they evaluate leaders. So that part of the work was made very clear, you can’t have leadership without that follower leadership relationship where the follower agrees to commit effort and energy beyond what’s expected to that relationship. And I just said, effort beyond expectation, that was the transformational leadership idea.
At the title of Bernie Bass’s book, it was The Leadership and Performance Beyond Expectation. When you look at the major leadership theories over the last, I think we’re talking 30 years now, for example transformational leadership, which was about empowering followers. LMX,, leader member exchange quality, which was about creating a high quality relationship where followers are given more autonomy. Autonomy to do what? To make decisions, to make contributions, to get involved. Even servant leadership, which has come on strong, even though it was introduced in ’70s, come on strong over the last 10, 15 years, where the role of the leader is to help grow the followers and their leadership capacities.
So a lot of more recent leadership theories, I say recent, last 30 years, I guess that’s not so recent. But it’s really about fostering followers engaging in at least informal leadership, empowering them, giving them voice. So that means two different things. Followership skills do involve when to step up and when to engage in leadership activities. A leadership skill is how to foster that. So we talk a lot about the whole notion of psychological safety which was introduced by Amy Edmondson in 1998, and Google did a big study that was published about six, five or six years ago, touting the importance of psychological safety for innovation, and for contributions.
What that involves is the leader giving voice to subordinates in a way that’s safer that followers, it’s safe for them to make contributions. So I think we have a much more complex understanding of what followership is at least from the perspective of stepping into leadership activities at times, and creating a much more network approach to leadership. But I also want to jump off of some more of what Ron said. There are particular followership skills. So if you don’t see the followers are passive recipient of leader influence, how do you then look at the follower in terms of managing that relationship?
So when I’m working with executives I see the eyes light up when I use the term, it’s not my term, it’s in the literature, skills in managing up. So how do you manage your manager? When I get asked this question, I suggest they have that conversation that revolves around the question, or the comment, help me help you. My job is to help my supervisor, my boss, my leader, accomplish what they set out to do. How best can I do that? And let’s have that conversation. And let’s talk about the best ways that could happen. That’s that managing up skill.
So teaching followers and working with followers in terms of how they can manage up has a big part of this. And helping to construct, social psychologists talk about relationships being constructed, that’s such a big deal and that’s in the leadership literature too. How do you construct the leadership follower from the viewpoint of the follower? I think that’s a critical point that leaders need to be thinking about as well as followers, as well as several of us who are in the business of leadership development and fostering effective leadership.
Daniel Serfaty: Thank you very much, Ron and Steve, about bringing basically what theoretical concept and models of leadership and followership can turn into very practical advice about how to develop those skills, and that very symbiotic relationship between leaders and followers; one help the other be better, in a sense, in their role. That’s good, I’m going to come back as a way to conclude in a few minutes about this notion of practical advice. But I want us to engage in a little prospective thinking right now on two dimensions.
One is, we just are approaching the two year anniversary of the COVID pandemic here, at least in the United States and in the world. That has in a way, some people say transformed, some people say accelerated an existing change that was there rather than taking 20 years, now that change is taking two years, into the very nature of work, and the very nature frankly of leadership. We talked a little bit about that at the beginning, but specifically with respect to your own experience with executive coaching, with what it says in the literature, some companies really literally collapsed over the past two years, frank or turned into a poor version of themselves, and some other companies thrived.
And the question people talk about the resilience, rather processes existing in the company that make them thrive. But also one of those key dimension is the leadership. The leadership that knew how to understand their new role, their evolving role. From your perspective, what would you say are, specifically about dealing with what is international trauma, or world trauma, this pandemic, what would you say are one or two key leadership dimension that got really brought to the fore, and also need to be developed in future leaders as we go beyond just the pandemic?
Stephen Zaccaro: One of the key points is about agility and being agile and adapting. COVID really hammered down several of the points we were making. The situation changed dramatically, literally in March, 2020 and in the academic world we had to jump immediately from in person classes to online. And for some of us who’ve been teaching over 40 years, that was an adjustment. We had to be agile in terms of thinking about how to do that. That’s the same thing with leaders. So they have to be agile in terms of, how can I achieve the core purpose and mission of my collective, my unit, my team, my organization, in this change context?
And the leaders who were successful were those who were innovative in thinking about that. So first we saw the need to immediately shift how you would do leadership. There were also more things we had to deal with. So I noticed the leaders around me were much more concerned about trying to help deal with the stress that people were experiencing, and trying to be more empathetic, and fundamentally changing how we were interacting with each other, interacting with our students in an academic setting, to manage that increased stress and so forth. That just continues to this day.
The distancing element, so much work happening in distance virtually, how do you manage that as a leader? And in one of my classes I’ve asked my students in discussion posts at a virtual board, how it changed for them and how they distancing, and how they had their leadership had to change. And really it was about trying to foster that leader relationship over virtual modes now. We’ve been talking about eLeadership or virtual leadership since the late ’90s, early [inaudible 01:03:58] but the COVID really hammered in for everyone, whether it’s a preference or not.
So what we talked about, being humble to adapt and the need to understand what situations require and adjust accordingly, that became so much more paramount and urgent in the COVID environment. And I feel like I’m just touching this surface of this so let me stop and Ron, if you don’t jump in and I’ll come back in.
Ronald Piccolo: There’s a couple very specific things that I think are heightened in the new world as it relates to leadership. One, Steve, you and I talked about this in the past where, as people are separated, whether it’s social distancing or interacting remotely, just as we are today in two dimensions, if you will, I think it heightens the importance for leaders to create connections among their team members or their group. Figuring out ways to foster cohesion and understanding and appreciation for each other.
Ronald Piccolo: We’re all kind of in a fast paced world and we kind of live like this. And so, this has always been a responsibility of leaders, to develop the interpersonal and inter team dynamic, but I think that’s as important, if not more important now than maybe it was two or three years ago, when people would naturally bump into each other in the workplace. A lot of that cohesion would just naturally happen, but it doesn’t happen quite as often anymore.
Ronald Piccolo: And so I think this could highlight the importance of the leader’s intentional effort to create relationships among the people who have to work together. That’s one key thing there, I think. Another key thing is just my own observation, with all the information that’s available to us coming from us with many different sources, and I think a justified skepticism about the credibility of a lot of the things that we see based on where it comes, there’s a ton of complexity here.
Ronald Piccolo: And one of the key parts of effective leader is to help navigate that complexity, and to help simplify it and to convey confidence, or create some clarity where there isn’t some immediately, or where people can get confused. And this is tempered against the desire all of us would have for leaders to be flexible and open and responsive. But on the other side of that, there’s value in the leader of being clear, being succinct, being direct and allowing people to have some understanding and clarity, and what can I believe in and what can I not?
Ronald Piccolo: And what are we going to do here? And how are we going to deal with this? I think the uncertainty about the impact of the pandemic and what it means, and are we safe, and are we going to live? That just creates a kind of anxiety and stress among people that could be mitigated by someone being clear and direct and firm. At least allowing the leader to play that function. To clarify where there’s uncertainty.
Stephen Zaccaro: Yeah I want to jump on that, I have excellent points. When we talk about flexibility, we mean flexibility and behavior, not flexibility and values, not flexibility in core functions. And I think that’s what Ron is alluded to there in part. I’m thinking back to the lead around me, who I thought were highly effective, my department chair, for example, some of the other leaders at George Mason University, where I’m at. As the anxiety ramped up, as stress ramped up, they would provide information.
They would offer clear direction on the key things we had to do, but they would also be reaching out with stress management saying, look, I understand what you’re going through, and I understand, even though they were going through at a much higher level, because they were excess of everything that was going on, saying to every one of their followers, “I understand please reach out to me if it gets too much.” They were dealing with the relational part of this, the emotional part of this, as well as the task part of this. This is what we need to do and here’s the ways we can do it.
There’s also was a lot of attempts to put the social part of it. How can we create that connection through this screen? And I will tell you honestly, I don’t know if we’ve fully gotten that. I know that because as people are now reconnecting they’re saying, “We miss this, we need this.” I think in the future we’re going to have to figure out the balance between the gains we get from the distance part of this, I think there’s a lot of positives with Zoom, for example, and the negative aspects I will give by way of example.
We conduct a learning series that are our university that was always face to face, we went virtual because we had to and realized we had a much better reach. So we’re going to keep that. But then how do we foster the connection among the students and faculty that that previous version had? So it just strikes me that the good leaders became more flexible in terms of the different things they were responding to. They needed to be task oriented to provide direction, clear guidance that Ron talked about. But they also needed to be attuned to the emotional load that people were experiencing.
And even if it was not their nature to reach out, so I understand that. And also the social connectedness that was beginning to fray. That sounds like a lot, that sounds complex, but that’s where the leadership skill comes in. The ability to recognize that that was needed just by my department chair reaching out and saying, “I know what you’re feeling, and please reach out to me.” That was enough for many of us to help manage that a little bit better.
Daniel Serfaty: Listening to you giving almost this image of not just how we coped in the past, we are coping with COVID, but also an image of the future. In fact I realized that many of those qualities that the situation forced on the leader, whether it was natural for certain leaders or for them, they really had to make an effort, were things that you described before about the whole notion of social influence and helping people, being there for them knowing that they are there for them. All these basic qualities are basically … have been emphasized by COVID.
So as a way to conclude, there are some people in the audience right now that are very, probably impressed, two professors who know a lot about leadership, and they are asking themselves, if they are either in college or they’re at the very beginning of their career, they’re seeking advice. If they were to ask you, I really love that idea of a leader, I’m not currently a leader in all the definition and the complexity that you have described over the past hour, what’s your advice about how I can develop my skill, I can put myself in the path to become a leader? What’s some practical advice you have for that 20-something year old listener who is trying to become a leader? Or to become a better leader.
Stephen Zaccaro: So first I would start by reflecting upon what leadership means to them. Do they see leadership as a power thing where I could tell other people what to do? Or do they understand the responsibility of leadership? What is leadership to them? Are they motivated to be leaders? So first, let’s be curious about your approach to leadership or why you want to be a leader, and also assess yourself as a potential leader. What are your strengths? That’s important. That you can apply to leadership. And then look for experiences.
his goes back to what Ron said earlier about the role of experience in development. Look for experiences in which you can start to engage in at least informal leadership, leader project. Step up to take over a team informally, and then look for feedback. Look for mentors, look for excellent leaders. And I often ask my students, who do you admire as a leader? Watch them. What are they doing that makes you admire them? Why are they effective? So find learning partners in the terms of mentors, models, or coaches, or bosses, or others that exemplify what you want to be.
So lay that foundation for understanding leadership and growing those initial skills by seeking projects in which you can employ your strengths as you understand them and stress them so that you can grow more and more leadership skills. That’s how I would start it if you’re aspiring a [inaudible 01:12:10] leader, start with that awareness, identify your strengths, look for experiences in which you can employ those strengths. Because if you’re successful in that you begin to build up what we researchers call leadership self-efficacy, the confidence that you can be a leader, and that would help you strive from even more leadership challenges, and that starts the journey.
And I love for Ron to step in, but I’ll say, well that’s the aspiring leader. Later in the career there are other things you can do, but I’ll stop that.
Ronald Piccolo: Yeah I think probably an honest self appraisal. For the inspiring leader and even for someone who’s in a managerial role or, maybe has had 20 some years of leadership experience, a humble, honest self appraisal of what you are good at, what you’re naturally good at, that could be setting a vision, how you solve problems, the manner in which you communicate, how you develop rapport with others, all of those things and embrace it. Embrace your strengths and identify how you can make amends or work around those things at which you’re not naturally skilled, I would say that for anyone.
Ronald Piccolo: Just start with who you are and let’s be comfortable with it and go from there. And I also think that is liberating in some way. Because if you try to be like de’ Mici or you try to be like Abraham Lincoln or Margaret Thatcher, it’s like, yeah, I don’t know that I can actually do that. So what can I do? And how can I be the best possible person of influence given where we are and what my natural instincts might be? I would start there.
Daniel Serfaty: Thank you so much, professor Zaccaro and professor Piccolo, Stephen, Ron, for fascinating hour about leadership. Thank you for our listeners for being with us today. This is a kickoff of a four part series on leadership, and that gave us a wonderful basis to understand leadership in practice in our future episodes. Thank you for listening to MINDWORKS, this is Daniel Serfaty. Please join me again for the next episode. We welcome your comments and feedback, as well as your suggestions for future topics and guests. We love to hear from you. You can tweet us @MINDWORKSpodcast or email us at MINDWORKSpodcast@gmail.com.
MINDWORKS is a production of Aptima Incorporated. My executive producer is Ms. Debra McNeely and my audio editor is Ms. Lindsay Howland. To learn more and to find links mentioned in this episode, please visit aptima.com/MINDWORKS. Thank you.
Daniel Serfaty: Welcome to MINDWORKS. This is your host Daniel Serfaty. Leadership. In the realm of human performance, there is no other dimension that has generated more theories, constructs books, controversies seminars than leadership. What is it about leadership that fascinates us so much. MINDWORKS decided to get to the bottom or the top in this case of it. And this is a second in a series of five episodes on the practice and theory of leadership in the 21st century. An institution that has embraced and developed the quality of leadership for centuries is a United States military. It would be a challenge to find more qualified experts than my three guests today, who will guide us as we explore lessons learned from the military and their impact on leadership in everyday life.
Scott Flanagan is a retired us army master sergeant, having served 20 years of active duty in special forces and special mission unit assignments. Scott’s army career included, advisor, operational instructor and research and developments assignments, including multiple combat developments worldwide. Over the past decade, he has been supporting multiple department of defense research agencies in the area of training leader development, assessment and technology development.
Dr. Fred Diedrich holds a PhD in cognitive science from Brown university and is currently a senior independent consultant. His work focuses on methods of instruction and assessment designed to deliberately support development in areas such as social skills, character, initiative, empathy, critical thinking, and leadership. And really directly relevant to today’s podcast, the creation of assessment tools for leaders attributes in the US army. Previously, Fred was president of Aptima and chief executive officer of Milcord.
My third guest is Morgan Darwin. Over the past couple of decades, Morgan has assumed the leadership of several global companies. He’s a founder of Sophia Speira, the services and technology company that develops and implements solutions and cloud wise platforms designed to measure analyzed and manage complexity in healthcare, workforce development and fleet management through his brand, such Sophia Solutions, Sophia Health Systems and Red Hill Mobility. Morgan co-led an US army initiative to redesign military training and education, to prepare service members for operations that merge combat with civil and humanitarian operations. He’s a 22 year veteran of special in the US military and has spent much of his adult life working in Africa, Asia and Europe on large infrastructure development projects.
Gentlemen, welcome to MINDWORKS. Let us start perhaps by introducing yourself, but also I would like to share with our audience your own experiences as a leader and what made you choose this domain as a field of interest? So let us start with you, Scott.
Scott Flanagan: Thanks for the nice introduction. My experiences in leadership compared to the other folks on the phone here, they’re not unique, but they’re a little different. I spent most of my time at a very small unit level. My experiences were at a tactical level in the military. I didn’t progress any much further than that, just because of the units I was assigned to and the jobs that I had. And so it was mostly small teams. I did hold a leadership position at a brigade level, which was research and development, which really opened my eyes to some of the unique aspects of leadership at a higher level, mostly in terms of developing and growing relationships with others. I’ve always felt like that was probably one of the most important things that I learned in the military.
As a civilian, again, mostly in leadership positions with small teams, training leader development, program management, and those kinds of things. I’m most interested in it because my experiences in the military really opened my eyes to how those things I learned about leadership can transition to places outside the military. I’ve learned from it and I always felt like we could give back to not only other military personnel, but civilians as well.
Daniel Serfaty: Thank you. We will explore that the correspondence between in leadership in the military context and in the civilian context a little later in the podcast. But Morgan, you also come from that tradition of military leadership, but you moved on after that and went on to manage global corporations. Tell us a little bit about your experience as a leader, but also what interests you into understanding and developing leadership.
Morgan Darwin: Thank you, Daniel. I really appreciate the opportunity to be part of this podcast. Leadership is something that is very, very important to me, but I would say that it’s something that is still a learning process. I mean, you mentioned the military part and also the civilian part and what I would probably say that is most significant to me about leadership, particularly in the opportunities that I’ve had to work in, is understanding the role of trying to provide leadership in a distributed environment. So whether that is in a military organization that has units conducting operations over a large geographical area, or as you mentioned with a global company, leadership of different business units in different countries or different regions, side of a country. The challenges associated with, how do you influence the team and how do you create an environment that fosters accountability for those subordinate leaders to be able to accomplish their mission.
And I think that what really interests me about this and I know that we’ll talk about this in much more detail is the emerging role of technology in leadership. And I think that technology enhances leadership, but it doesn’t replace leadership. So I think that this idea, which is particularly of interest to me of distributed leadership, is how do you build the relationship between team members, between the leadership of different teams in this distributed environment in a way that allows there to be more efficient execution of either the business function or the military function? That’s my kind of interest on this particular subject.
Daniel Serfaty: Well, thank you. I’m eager to dig deeper in that direction because as you know, this is also an interest of mine. And Fred, you come from a different dimensions. You are a bonafide psychologist with a degree from Brown university. You could have chosen different directions, child development, which I know you focused on early in your career. What fascinates you in particular about this notion of leader development?
Fred Diedrich: Thank you, Danielle. And likewise, thank you for the nice introduction. I think you just referenced developmental psychology and that was the work that I did largely in the university in the late ’90s. And I think that’s the answer to your question about my interest in leaders, in a particular leader development. And so as I think that we get into the work with Morgan and Scott, we’ll hear a little bit about kind of the history of what someone has done mattering in terms of how they behave now.
And so in other words history matters, in the case of leaders, we need to think a little bit about what’s going on early in careers because that will shape what they’re capable of doing or not later in careers. And so the short version of all that is that my interest in leadership is really one that fits with an interest in human development. And I think that many of the lessons that we know from even early child development turn out to be the same kinds of things we might worry about with developing soldiers and sailors.
Daniel Serfaty: That’s an interesting, not just an analogy, you’re making a scientific case that some of the processes are the same. We’ll explore that. But I have a question for you, Morgan, which is really key. In today’s highly connected world, totally networked, where you can have access to anybody at any time, automation, intelligent devices, are leaders less important than before or more important than before? In a sense, is technology lessening the impact of leadership or enhancing it?
Morgan Darwin: So I definitely think it’s enhancing it. I think that technology allows us to do more and expands the domain that you’re able to influence. In the past leadership was primarily executed through the people that you could directly influence that were within eyesight of you. But now we have the ability to have teams that are literally scattered around the world that are all connected. And so I think that all technologies really are designed to enhance the capability of humans. I think that it may create a requirement for certain attributes of leadership.
For example, you still have to build trust. You still have to build accountability. You still have to grow judgment and problem solving in your team. You may have to think about it differently if you’re not physically on site and you’re trying to influence it remotely, but to me, technology is never a replacement for human beings or the relationship that human beings play. And particularly with AI, I think that you, in the future you could see AI inserted into the leadership structure, but it’s not replacing the relationship there. And in many ways, AI has to have very similar attributes as the individual humans that are in that leadership team relationship.
Daniel Serfaty: Fred, do you agree with Morgan on that?
Fred Diedrich: Yeah, I do, but I don’t think it’s a matter of are leaders relevant are not. I think that consistent with what Morgan said, I think that the environment and the needs for leaders are different. So in the case of a distributed team like Morgan is outlining, he suggested several kinds of attributes. I might add to that things like character and initiative, where those attributes become, I think increasingly critical as you are distributed and operating independently. But I would characterize that as being a different set maybe than they were in some times in the past, rather than are humans mattering lesser or more, or technology making leadership less or more relevant. The real question is how it changes.
Daniel Serfaty: Scott, do you want to chime in on that?
Scott Flanagan: I was thinking about, in terms of some of the research that I’m involved with now, where machine learning is making certain tasks much easier to do. And I would argue that they’re accurate as well, but they’re only as accurate as the information that we have sort of put in under the hood. That’s where I thought leadership really needs to continue to be on the forefront. With what Morgan and Fred said too, there are people sitting around conference rooms having discussions about what’s important and what’s not. Prioritizing, defining, all of these things that will go into these systems, and I think leadership plays a primary role there. The garbage and garbage out idea. And so leadership plays a critical role at the beginning all throughout, but at the beginning, for sure.
Daniel Serfaty: We know very well that there are as many books about leadership than there are leaders these days, and each one of them has a four dimension of these, so the seven layers of that. But you are an experienced leader in the field, what do you think is the number one or two attribute that leaders must have no matter what?
Scott Flanagan: For me it’s, I have found over the years, especially in the military, but even outside the military, trust between subordinate and leader, and leader and subordinate and encouragement. And the best leaders have always encouraged me. They trusted me and they encouraged me. And gave me the autonomy me and sometimes the authority to do things, even when I thought that’s just too much. That’s just too much responsibility. I’m not prepared for this quite yet, or at least I don’t feel like I am. But they trusted me to do it anyway and they encouraged me to do it. And so those two things for me personally have been probably the most important attributes or competencies that I relate to leadership.
Daniel Serfaty: Trust and encouragement. Morgan, which are your top two?
Morgan Darwin: So there’s a long list of desired attributes on both sides of that. But then [inaudible 00:12:46] boils down to me is being the most important is a sense of empathy, to be able to understand what the team is going through as individuals or as a group. And then the second attribute is the ability to inspire the creation of value. And I think, the creation of value is not just on the commercial side, it’s on both the military and the commercial side. It’s creating the value that the organization is formed to create and the ability to inspire that, and the empathy that goes along with that, to me are probably top of the list.
Daniel Serfaty: That’s interesting. And you believe they cut across domain.
Morgan Darwin: I do. And I think that one of the things that I think is sometimes missing from idealized conversations about leadership, is the fact that in the end the organization has to create value to justify why it exists, why it was formed, what is purpose is. And so we can never forget that. While we want to have leadership that is encouraging and growing at a developing subordinates, at the same time we have to find a practical balance between over-idealizing a perfect leadership environment versus effectively creating value for the organization.
Daniel Serfaty: Cool, interesting. This notion of value. Fred, do you want to chime it from your perspective again. And again, looking at both the leaders that you helped develop say in the different military schools, but also your own experience as resident or a CEO of a corporation.
Fred Diedrich: Yeah, sure. And I will steal a little bit from both Morgan and Scott. So, two things stood out, Scott called out trust, the ability to build trust within the team. And Morgan called out empathy. And part of my answer to this question would be character, but expanding on that for a moment. It’s worth noting that the army has done an incredible job, essentially developing lists of what they want out of a leader. And the current list has say 25 or 30 attributes. And you can probably find lists with more if you want, right? But I think Scott was very wise in picking trust or what the army would say is builds trust. Because when you think about what it takes to build trust amongst the team, many things matter.
So Morgan talked about a dynamic balance between achievement and developing subordinates or whatever other factors might come into play. So when we think about it from an army perspective, we might have things like achieves, that’s one of the things the army wants out of its leaders, to be able to achieve mission first. And if you don’t do that, you’re not going to build trust. If you’re not a successful leader, you will not build trust. At the same time though, the army tells us that we need interpersonal tact. The army tells us that we need empathy. The army tells us that we need character and I can go on, and all of those act in concert to build trust and maintain that dynamic balance, which is at the heart, I believe are what both more and Scott talked about.
And I’d add that I think that’s really hard. And oftentimes it’s underestimated just how hard it is because I think it is dynamic. It changes over time. There are times when you have to make a trade off between one element or another. And when you make that trade off between one element or another, you effectively make a withdrawal on trust sometimes, right? When you do that. And you have to. And Scott, I’m stealing that kind of withdrawal on trust from your checking account idea from you, because I know I’ve heard you talk about it in much of your work that you did over the years with special forces.
Daniel Serfaty: No, it’s fascinating that those components, the reinforcement of Morgan’s answer to my previous questions about to which leadership is still relevant. Because these are almost like meta qualities that go above your ability to make decisions and your ability to manage uncertainty, and your ability to do all these transactional things that leaders do, which to a great degree can be supported and sometime even replaced by some automated machines or some networks or some other things.
But those qualities above that layer, the one that you just mentioned, empathy and trust and building values and trading off, as you just said Fred, are things that we are far from having any technologies that approximates that. And I think perhaps that’s what the leaders will do in the future is move up to that next level of worrying about or focusing on those qualities rather than the transactional aspect of leadership perhaps. Just a hypothesis. Talking about that, there is an old myth in the literature, you can still find it in some unfortunate books in business schools. There is a myth that true leaders are born, not made. All you do at the end is just shave a little bit the corner. Do you agree or disagree?
Morgan Darwin: Okay. Well, what I feel is that there’s a long list of attributes that define a good leader. And based on the prioritization of a unit or an organization, what they’re looking for in leadership might vary. And so we may look back at how history and say, some people were great leaders. Well they may have been great leaders achieving certain results, but they may not have been great leaders as it relates to mentoring and developing future leaders.
So I think that when you look at all these different attributes, where you begin to recognize is that every human being we’re all born on a continuum. And the mix up of those attributes is kind of like how the cards are dealt you. And so what I think is that, some people maybe they are born a little bit further to the right on that spectrum, some people are born to the left. But our experiences and how we’re mentored over life can push us in a positive direction or in a negative direction. So I think that everybody has the opportunity to learn and become a better leader regardless of where they originally are positioned on that spectrum.
Daniel Serfaty: That’s very comforting because I’m sure that many… People in our audience are wondering, leadership is for other people I’m just engineer or a good worker and leadership is not for me. And you’re saying that, “No, that’s not true. No matter where you stop on that axis, you can always strive towards leadership qualities.”
Morgan Darwin: Yeah. And remembering that leadership is not always the position that you hold. Leadership is doing something well. And so it doesn’t really matter what your position is in organization, there’s leadership opportunities that are relevant to that level of responsibility. So I think that everyone is a leader and then there’s opportunity for all of us to grow as leaders.
Fred Diedrich: I really agree by the way, with the point that Morgan makes that leadership can come in many forms at many levels and there are opportunities to grow. So, I deeply believe that humans can grow in any number of ways, including as leaders. With respect to this question about whether or not leaders are born or created, I would simply add that it is no doubt in fact, the wrong question. And it’s the wrong question because it fits with the nature, nurture debate, which has gone on for hundreds of years. Do we come with stuff or do people arrive as blank slates?
And the fact is that everything we do, every interaction we have in life is going to be determined in part by both. And I have no reason to believe that we can’t develop leaders in multiple dimensions, especially when you think, as Morgan noted, how many dimensions there are that make up a good leader. It’s complex. You’re not talking about developing one thing, you are talking about developing a lot of things.
Daniel Serfaty: Thank you. Scott, to go there, maybe based upon what we just heard from Fred and Morgan, could you share some anecdotes without getting into operational details of obviously, but some anecdotes in your own experience, even as a soldier, as a junior soldier, when you started your carrier in the military, when you saw someone and you say, “Wow, that’s a great leader.” What did you see in that person? Can you tell us why in the eye of 20 year old suddenly that person was a leader?
Scott Flanagan: Yes. I have someone in my mind right now. It was a little later in my career, I think I was a late bloomer, but I could tell there was something, there was like this aura about this person that really struck me as this person’s going to be a senior leader of this organization, I have no doubt. And I thought about him quite often. As it turns out he was. I saw him go through the ranks and become a very senior leader in the organization that I was in. He went on beyond that and continued his leadership role at higher and higher levels as a non-commissioned officer at probably some of the highest levels in the army. What stuck out to me about him was that he seemed untouchable in terms of he wasn’t arrogant or anything like that at, but he seemed to sort of just rise above the… Sometimes there’s nonsense in an organization or gossip or frivolous kinds of things. And he seemed to stay focused on being a good person to everyone around him.
He was a very high performer by the way, sort of a top performer, but he didn’t sort of blog that over people. He just was. He was very encouraging to those outside of his own little team or his own little troop or whatever he was in. He was always encouraging to others outside of that small circle. And when he was in a leadership position, he was very competent. He was good to those around him, just generally speaking. I never really saw him go sort of into that, talking about people and negativity and all those things that can happen in an organization. That’s what stuck out to me.
Daniel Serfaty: I wonder whether or not a combination of this behavior that you describe, we don’t know this gentleman. I don’t know this gentleman, maybe Fred and Morgan do, but instinctively, it inspired trust in you, the soldier, that basically you’ll follow that leader because of all this quality of discretion of being even killed and other things like that. I wonder Fred and Morgan, if you have such an example of especially looking up early in your career, whether it’s in the corporate world or in industry, in academia, in the military, when you saw someone and you sensed that, that person was a great leader.
Morgan Darwin: So Daniel, I can share experience. When I was a very, very young soldier shortly joined the military. Now in reflection on this, this is not something that I fully realized or understood. You learn more about leadership, particularly once I met Fred and a couple of other people after I retired and we began to work with the army on leadership. But in this particular situation, I was a brand new private in a company, and my company commander who had no business of knowing who I was, I always remembered thinking that there’s nothing I wouldn’t do for him as related to the mission. No matter what he asked me to do, no matter how dangerous it was or how difficult I was going to give my all.
Several years later was now a mid-level MCO, he was a senior officer and we crossed paths. And the first thing he said to me was, “Morgan, how’s it going? Is your wife still a school teacher?” And in reflection, what I realized is that my company commander knew that as a private in his company my first name was Morgan and my wife was a school teacher. And even if he asked me to do something that would endanger my life or was extremely dangerous or difficult, he was doing it knowing who I was as a person. And that has stuck with me all of my life and trying to understand that these are real people in our organizations. And we may have to ask them to do really difficult things or challenging things, but we should never forget who they are.
Daniel Serfaty: Maybe that’s why you emphasize empathy quite a bit in your explanation of leadership. That’s very good. Fred, any example you care sharing with us?
Fred Diedrich: Sure. I’ll go way back in my career, but I’m going to twist it a little bit to ask the question and say, what kind of leader do you not want to be? Because I think that can be equally influential. Before I knew anybody on this phone call, I worked for a different company and there was a member of executive management that was sort of exceedingly hostile and irritable and kind of demeaning. And he had a long track record of really creating a non-productive environment in terms of enabling his employees to feel good about themselves and feel good at about the work they were doing. And the consequence of that was the unit that I worked with I believe turned over by about 90% in two to three years. And that’s pretty devastating. And you’re not talking about a turnover of people who have skills that are easily interchangeable. Because everybody, I just mentioned had a PhD in some kind of engineering.
So, that’s another example. And I think that seeing both the positive and the negative can make a huge difference in terms of those you might model or not. And I believe that, Scott earlier brought up the idea of modeling, when he mentioned sort of some of his role as a consultant. And I think that is really super important. There are formalized leader training programs. But probably what you learn more than anything else is from the people you work with on a day to day basis. And sometimes what you talk about when you go out to dinner or you’re riding from the airport down to the client site or whatever it may be, as being profoundly important in terms of how you develop as a leader.
Daniel Serfaty: Well, thanks for sharing that Fred because I think it’s important we learn both from those positive and negative examples. And sometime we learn instinctively, but sometime also we reason about that and we apply those lessons later. Scott and Morgan, I hope you were relieved to know that none of the three of us were involved into that anecdote that Fred just shared with us [inaudible 00:27:24]. Since we all worked with Fred at one point or another.
And I want to go back to that actually. The three of you, I remember more than a decade ago, started to work on a project to really change the way the United States army was thinking about training. And I remember, me being a colleague of you Fred, that you were really excited about that project. You thought that was really a game changing project for you as senior researchers, but also for other folks. I would love for you to share that with our audience, to the extent possible. What was the challenge at the time and what did the army want to do with that kind of changing the mindset in terms of training?
Fred Diedrich: Yeah. Thank you, Daniel. And the project you’re referencing is, I think still to this day, the most important project that I’ve been involved in over the last 20 years, although I will note that it goes on in many ways. Most of what Scott and I in particular do on a day to day basis has something still to do with this effort, 20 years later. But I think I can better speak to what we did and why we did what we did, but I’d like to defer to Scott or Morgan first to characterize what the operational problem was. And I think that what you will hear is that, that is a starting point and it has to be the starting point for everything that comes after. So Morgan or Scott, I’m not sure which one you would like to kind of talk about the genesis of the problem, but I think that would really help.
Daniel Serfaty: I think this is very important because for those of you in the audience that are not familiar with the military, a lot of major behavioral science progress, whether it’s on leadership or decision making or team dynamics, was actually started because of a need or a challenge that the military thought that they had and eventually propagated to every single aspect of our lives. The example that Fred described is certainly one of them. So even if you’re not very familiar with the operations, those behavioral science issues are actually very pervasive in every aspect of our lives and our world. So yes, Morgan or Scott.
Scott Flanagan: I’ll start it off. I’d like to first say that the Army is the best army, I think personally in the world. There’s always room for improvement, but the leaders in the rank and final soldier in our Army are amazing. What I had the privilege to learn being in the special operations community it changed my life actually. To be in an organization that valued my opinion to a fault, to be in an organization that respected and trusted me to do things that I never thought I could or would do, to engage me in conversations, meaningful conversations at the lowest level of the organization. And value my opinion to the point where I had to pause oftentimes before I spoke and really think deeply about the responses, things they were asking me to do, they really made me feel not so much important, but humbled in a way.
And so I really had to think critically and be more introspective and over time, that was just continually encouraged there. When I was taught or trained in that organization, I felt like I was an integral member of the team from the very beginning. They spoke with me just in a very professional, meaningful way. I don’t know how else to describe it. I felt like after I retired that some of those approaches to training and leadership may have value with the rest of the army, the conventional force in the army. And I was in a position where I could potentially influence that. We, as a company at the time, we were in a position to go around to the army and work with many, many units, briefings, training discussions, and it wound up being that we had an opportunity to actually do some training for some of the forces that were getting ready to go to Iraq and Afghanistan.
And so we built a small team of people and we started doing some training courses for them, but the gist of the training courses was not so much to teach them a skill, although we didn’t pay lip service to the skill as was stated sort of earlier, in the army performance matters. It matters in the commercial sector too. We have to perform. Fred as a scientist has to perform the scientist role and so we have to do that. But there’s also another aspect of it which is problem solving, accountability, critical thinking, et cetera. That happens all the time in the military, especially in combat situations. And so we thought to emphasize that in our training, our approach.
My philosophy was, if you build it, they will come. If we just demonstrate this and model it, they will get it. And I’ll quickly sort of segue into Morgan because he had a profound impact on thinking a little bit deeper about our messaging and what we really needed to say to them in terms of developing these competencies and attributes, because we thought that those were just as, or more important than the skill itself. We can develop hard skills, but if we don’t do it in a manner that develops the other competencies and attributes, then we’re not meeting a standard that we think is important.
Fred Diedrich: Yes. So picking up on this a little bit, these other attributes, right? So I think that the genius of this particular program, I’m talking about the genius of people like Morgan and others that were involved in it, was a realization that, if in fact we are interested in things like critical thinking, problem solving the various elements that Scott just mentioned. And that we want those in our leaders. We had better look back into how these individuals were trained, really from the time they entered the army. Because that’s going to set up a pattern of behavior that’s reinforced over time or not. And that will be where the answer lies. So I can pass it back to Morgan or Scott, if you want to expand on that at all. But I think that the key insight is that you don’t want to wait until 10 years into somebody’s career when they’re taking on a significant leadership position to start to worry about these attributes that you want at that point in time.
Daniel Serfaty: That’s very interesting. And do we believe that we succeeded that early injection of those skills that are very complex skills and attitudes, larger skills, actually changed the trajectory of those soldiers? Morgan.
Morgan Darwin: So I absolutely believe that it did. And one of the key things that Scott kind of identified was that, particularly in the very beginning we weren’t really sure what we were doing other than trying to demonstrate good leadership and training. And obviously good leadership is more than just about knowledge transfer, it’s also about the growth and development of the people that are participating in the activities. And what we began to do is during the course of the training during the day, we would be talking about the task and we would be talking about managing stress and the performance task and mass train the basics. But then at the end of the day, we would do an after action report where we would discuss with the students, what they saw us doing and how it impacted them in the training.
And what was really interesting about this is that it’s a lived experience, hence it meant something to them. Even in a five day course, we would actually have people come up to us and say that in my entire military career I’ve never had anyone explain anything to me before or value my input into solving the problem. And so we would see profound changes just in the five days that we would be working with these mid-level leaders. And so I think that kind of the argument of what Fred is making and what Fred has got to continue to work on today, is that if you introduce this concept into the initial training of soldiers and officers, that it has an accumulative effect over time and becomes deeper and deeper ingrained in them, they expect as a soldier to be a valued member of the team. They grow up with the idea that they are expected to perform, they’re expected to be trusted. All of these different things become a component of that. I think that that’s really, really important.
And I would also add that, Scott and I were coming out of military background where we were just demonstrating how to a train based on our experiences, or maybe even intuitively. And it was the opportunity to work with Fred and other people like Gary [inaudible 00:36:22] they’re basically a scientist. They said, “What are you trying to accomplish? What are the metrics that will allow us to measure that? And now let’s go out and look at it.” And so it really became a upward spiral of understanding for all of us and realizing that how you teach and how you train actually has a profound impact on the development of these leadership skills.
Scott Flanagan: It had a great impact on me to working with the science community in that area. To this day, I’m still working with scientists to this day because of that. And how Fred and team had a profound impact on helping us define what it is we’re doing like Morgan said, and then putting some rigor behind assessing it. Even understanding at the time what assessment meant. How do we assess confidence or accountability or problem solving? We didn’t really have that sort of experience or expertise, but we learned from the science community how we can not only understand that more, but how it impacted how we implemented training also. So there’s those sort of give and take on both [inaudible 00:37:25] there.
Fred Diedrich: Yeah. Let me just add one quick thing to that and then back to you, Daniel. On the assessment side, and I think this may help clarify some of what we’re talking about here a little bit. So one of the results that be found and this was sort of 2008, but through different opportunities with different clients, it was replicated two or three times [inaudible 00:37:48] say 2015, was really very simple. And it sounds almost trivial, but I think it’s profoundly important. And that is if the teacher, or in this case the drill sergeant or the leader whoever it may be, gives the student, in other words, the soldier, the trainee, the opportunity to exhibit a skill, generally speaking, they will do it. And if in fact you do not give people that opportunity, generally speaking, they will not.
And we found that through correlations. And what I mean by that, if you ask a very young soldier someone who’s been in the army for a couple of weeks to solve a problem, you will see behaviors in that soldier trying to solve the problem. Doesn’t mean they’re going to solve it. That’s irrelevant whether or not they actually solve it, right? Then the question is whether or not they engage in those activities. And that’s really important because that sets up the habit. And I don’t necessarily mean habit in a sense that psychologist talk about habit, but it is kind of a habit of showing initiative, a habit of solving problems, a habit of critical thinking that matters and will be repeated over time. But if you don’t build that habit early, you will not see it later.
And part of that equates by the way to believing, and this goes back to trust. So Morgan and Scott articulated some of those earlier, but it’s trusting in the young soldier that they will in fact rise to the occasion and that you can trust them to engage. And that matters profoundly because that’s really what gets you out of the starting blocks into this accumulation over time, as opposed to an environment which assumes that this young person is not capable, therefore I cannot trust them. What Morgan and Scott did and their colleagues was to introduce a paradigm, which said, “I’m going to trust you. Within bounded conditions, so I’m not going to let you fail, but within bounded conditions I’m going to trust you. And then we are going to trust in the human organism to want to solve problems and want to grow.”
Morgan Darwin: Fred, just tying into that. One of the things that we encountered, and Scott jump in on this is how profound the idea existed across leadership that the brand new soldier or the young soldier could not be trusted. They needed somebody to guide them through it. They needed somebody to walk them through or explain to them what they’re supposed to do step by step. And we actually counter significant arguments that you could trust a young person to solve problems or rise to the occasion as you described it. But what was also interesting as we went through a five day course of training, and we exemplified that trust to the cadre, which were mid-level leaders, you would begin to see a significant number of them began to say, “Okay, I’m going to replicate this trust within the boundaries that you defined Fred. I’m going to this trust when I get back to my organization.”
You would even find some of them actually talking about replicating this trust with their own kids. But it’s very pervasive to not trust. It’s almost like it was a natural tendency to not trust the younger generation to be able to solve the problem. Primarily because you perceive that they don’t have the experience that you have so you naturally assume that they can’t solve it.
Daniel Serfaty: You can’t imagine how much I appreciate the sharing of those insights about that particular experience and how profound the impact you’ve observed on the soldiers and on the trainers, but also on yourselves too, which is fantastic. I think that in a sense, many people that are working in corporation today should listen to what we heard in the past 15 minutes here, because quite often in corporations, corporations just put a set of values on the wall. We value, I don’t know, integrity, this, that, and that. We value initiative. We value entrepreneurial spirit. But they do nothing in terms of developing their employees to reflect those values. What you created there, you kind of reverse a process through that process, you created the value that was then replicated throughout the organization. That value of trusting that folks can actually develop themselves.
I think many corporations today do not think of that. If they hire a young graduate with a degree in software engineering, they’ll put her or him in a cubicle and expect them to produce code. And they would think about developing that person maybe five years, maybe 10 years into their career so that they become department leaders or division managers or beyond, as opposed to instilling that notion that very early on, as Fred’s safe, if you don’t instill it very early on, you won’t get the result that you’re expect. I hope people take from that a beautiful lesson that is emerging from it. Thank you for sharing that by the way.
To continue one aspect has to do with leaders and followers. An aspect that is not very well studied in leadership studies is this notion of followership. Leadership followership. In a sense you all mentioned for leaders to succeed, they must have followers that trust them, that would carry on their intent. Accomplish the mission, so to speak. How does one develop followership skills? Is that something that we assume everybody has, or is that something that should be done intentionally?
Morgan Darwin: Very quickly? Daniel, I’ll go back to the story that I told about when I was a private and my company commander. He very well developed followership in me, I feel, but as you continue to walk through life with these different experiences of this desire to follow, you began to realize that followership is more than just following the good leader. Followership is really about accomplishing the purpose of the organization regardless of leadership. And so that creates an opportunity for you as the follower to recognize that I still have a role to play. I still have responsibilities. I still have problems to solve regardless of who the leadership is above me. And in the end, that’s the type of followership that you want to mentor so that the organization is moving forward regardless of leadership.
Scott Flanagan: I feel the same way. In my experience if I understand the organization’s higher purpose, that’s always motivated me to take initiative and follow. And it’s much like the research that I’m currently doing. If I understand the research question at my level as a nonscientist, then I can continue to help and assist and steer the research team towards those from my operational perspective. If I don’t know those things or I don’t buy into them then I’m not effective as a follower in that team saying, “I feel like I’ve been in a followership role most of my adult life. And I kind of like it there. I’ve been put in leadership positions many times.” But I think it’s almost a different topic, but I’m able to step into leadership roles that much easier as well. If I understand the purpose of the organization, if I understand the problems to solve, I understand my leadership’s concerns and what I can take off their plate. Then when I assume a leadership role, it’s that much easier for me to do that. And I’ve had to do that several times.
Daniel Serfaty: That’s a great hypothesis. God, it takes maybe a great follower to make a good leader. Perhaps this is a key to the notion of the empathetic leader that Morgan talked about. How would you understand how to motivate your followers unless you’ve been there yourself? That the kind of perspective taking can only come from having the discipline of being a good follower.
Fred Diedrich: And I think that’s right. But Morgan and Scott were articulating something I think very important but very subtle. So you’ll note that Scott did not articulate the idea as a follower, being somebody who will do what they’re told to do. Scott articulated the idea that he needs to understand the purpose of what he’s doing so that he can exercise the initiative to be able to get the job done that needs to get done. And that’s, I think, profoundly important. If we’re talking about followership and you expect somebody to follow you to do what they’re told, at Indiana I can remember one of my mentors driving into the parking lot and on the bumper sticker on her car was, “Challenge authority.” And so as graduate students, we were raised to challenge authority. But the thing about it is, is that the army, I think, has articulated the idea here very, very well.
And I’m not a soldier, so Scott and Morgan please correct me when I go awry here. But you inherent to the concept of mission command, which is the way that missions should be planned across levels, a higher level of the organization may specify the purpose of the operation. They may specify the end state that you’re supposed to reach and probably some key things that need to get done. But what you are not supposed to do is to tell your subordinate commander, how they will accomplish that sub task. And that strikes me as profoundly important because it is a doctrinal principle that enables the very initiative that Scott was articulating.
Daniel Serfaty: And we know the damage of not doing that, how detrimental it can be to commercial organization when you try to micromanage by telling people not only why they have to do what they do, what they have to do, but also how to do it. It’s a big of a pathological business process that it’s fixing. So this notion of management by objective in a sense, what they call it in business schools what you just described Fred is really key here to understand this followership leadership relationship. Which is very symbiotic, it’s not one directional that’s the whole idea.
We’ll be back in just a moment, stick around.
Hello, MINDWORKS listeners. This is Daniel Serfaty. Do you love MINDWORKS, but don’t have time to listen to an entire episode? Then we have a solution for you. MINDWORKS minis. Curated segments from the MINDWORKS podcast condensed to under 15 minutes each and designed to work with your busy schedule. You’ll find the minis along with full length episodes under MINDWORKS on Apple, Spotify, [inaudible 00:48:36] or wherever you get your podcast.
So we’ve discussed examples, wonderful examples you shared with us from domains in the military, in defense, in security about leadership and leadership development. And I wonder the degree to which good leaders are universal. Once you learn good leadership skill, whether it’s when you’re a boy scout or an army special forces, or the cyber operator somewhere, those skills are transportable to other domain, if you become the hospital manager, or you are on the manufacturing floor, or in the corporate suite. What do you think? Are those leadership skills transportable, or they are very specific to a particular domain and you have to start from scratch when you move to a new domain?
Morgan Darwin: So Daniel, I think that they transfer, but I think that you also have to recognize that the environment is very, very different. So real quick on the environment. In the military from day one, there is a very clear objective of services to shed each individual of a lot of their personal identifiable remembrances, and really build you as a team. And also in the military pay is pretty much standardized. And so once you embrace the purpose and the vision of the military and the mission, you’re not really thinking about compensation, it’s all about accomplishment of the mission. You have to be more practical on the commercial side and recognize that you can have a very inspirational vision for what you want to do with the company, but in the end, people have to get paid because they have expenses and they have requirements in their lives.
And so you’ve got a slightly different dynamic. But we are in the process of increasing the size of our company. And there’s one individual that I’m talking to that has very interesting history that transcends a military career as a reservist. His last command was brigade commander of a medical unit in the army reserves. He retired from that position and he is now a senior administrator with the VA. So he’s working on two sides of the fence. He’s got this military career that’s covered over 35 years of reserve and active duty and then he has the working on the commercial side, I would say. And when I’m talking to him about vision and team development and leadership, he blends them together as one thing.
So you may have to recognize that there’s a different environment that you’re operating in and the factors that influence and motivate the people that are working for you are different. To him, leadership and the development of people are the same. What motivates people may be slightly different, but treating them as human beings, involving them in the process, giving them responsibility, holding them accountable in a fair environment, he believes and I believe as well, works on both sides. That’s for the type of leadership attributes that we’re primarily talking on this podcast. The leadership style that you see growing in the US military over the past two decades to this individual it doesn’t seem to be any difference between commercial and military.
There’s different times of history in the US military where a different style of leadership was required. The military, for the past two decades has been very focused on growing and developing accountability, initiative, problem solving at the lowest level. That may not have been true 40 years ago. Doesn’t mean certain leaders weren’t doing that, but as an institution that may not have been true.
Daniel Serfaty: That’s interesting. Fred, Scott, do you want to chime in on this notion of transferability of leadership?
Scott Flanagan: Well, I’m just going to add quickly. When I look at the attributes and competencies that we’re familiar with in the army, I’ve always seen them as enduring outside of the military. I think I was thinking in terms of how maybe Morgan was thinking a little bit too, one of the biggest challenges I had coming from the military to the civilian sector was caring about the business side of things. I had a hard time sort of adapting to that because I just didn’t really like it. The military has its budgets and everything else. We have to be good stewards and understand those kinds of things, but getting into the nuances of those things it’s a challenge. But I do think that the competencies or attributes like adaptability and recognizing my roles and improving performance on the business side was something I had to learn.
Fred Diedrich: I think it’s the case that the context is going to matter, right? So there is something about the ability to lead others, interpersonal skills, things you might do to build trust or not, that are going to matter across domains. But how they matter in a given context is going to matter. And that’s true within the military as well as it is outside the military. And so I don’t think it’s as simple as moving from domain A to domain B. And if you’re a good leader, check the block. Because I think the mission and the context matter in terms of how that all comes together.
And I’d also note that if we go back to how the army defines what it wants from its leaders, it does include things like leads by example. Our stewards, the profession develops others. These are things that the army calls out. And it’s awful hard to develop others or lead by example if you don’t understand the domain that you are working in. So the kind of tension that I think Scott was articulating reflects that. It is the case that some of these skills will likely transfer, but there’s other things that probably have to be learned that are going to matter to make you effective.
Morgan Darwin: To build on that and to go back to the person that I was sharing their experiences about. Just so this is clear that we’re not talking about a feel good story, and tying into what Fred and Scott are saying, there has to be objectives and metrics on both sides. And so on the military side, the primary objective that he had in his style of leadership really is tied to readiness. And then the ability to effectively performed when call to mission. And we talked about those experiences pretty extensively. But what’s also interesting is that, that same philosophy that he was using to develop readiness on the military side, he gave some very significant examples of turning around organizations and dynamically growing customer satisfaction, dynamically growing, basically sales and customer volume.
And again, it all ties back to the development of the staff and he really ties turnaround time on work orders or customer satisfaction to the empowerment of the team and helping them understand what the vision is, what the mission is, and then giving them, or trusting them to solve the problems and giving them the authority to solve the problems. And so that has led to dynamic turnarounds on customer satisfaction, as I mentioned, which of course leads to greater volume of revenue and profits for the company. It’s not just like I say, a feel good story. There’s real evidence behind performance on both domains.
Daniel Serfaty: No, I agree. And perhaps these adjustments we talked about and context matter, there are some things though is true. I mean, it’s sometime within the same domain, the same industry, a leader in a particular company that is very successful, changes a culture and goes to a similar company but with a different culture and fails miserably. We have many examples of that. I wonder if certain qualities are absolute. We talked about things like trust and empathy. Maybe they take different forms in different cultures or different organizations, but they are there and they are kind of the necessary ingredient. They may not be the sufficient ingredient, but they’re the necessary one.
And I wonder above all that, whether this notion of ethical behavior, ethics, ethical rules is really actually above even those notion of empathy and perspective taking and trust and interpersonal relationships. What do you think? I mean, we see so many examples. Usually when we have a failure of leadership in the news, whether it’s political or organizational or corporate, there’s some kind of an ethical breach at some point, I wonder whether or not these are the things that one needs to emphasize above and beyond the transactional aspect of leadership or even the team building the interpersonal aspect of the leadership. Yes, Scott.
Scott Flanagan: When I think about ethical, I think about something that we learned in the army, or maybe broader than the army. We were given some autonomy to make decisions with the authority to do so as long as it was legal, moral, or ethical, that was just sort of a common phrase we would use. When I think about ethical today, it’s relatively the same as I thought about it then, that it relates to integrity. And integrity to me relates to trust, it’s a component of trust. And so I think if I was going to prioritize one over the others that you mentioned, ethical would I think would be at the top just because I think if we violate ethics, we’re violating trust and everything is downhill after that. So I do see it relating to trust.
Morgan Darwin: I think you can also say that critical aspect of ethics is, one, there’s very few secrets in any organization. So your behavior and your decision making process, if it compromises any of the values of the organization, people are going to know about it. And so in the end you could almost say that ethics are the values of the organization. So if you compromise those, then basically what you’ve done is you said there’s open season on those values for anybody. And your level of compromise is how somebody else perceives that you could be giving them the green light to do things that can really be harmful for the organization.
So from that perspective, whether it is taking a slightly unfair advantage of a client and a business scenario, or doing something that potentially is illegal or unethical, or would bring ill repute on the organization, anything inside of that range other people are going to see it. And they’re going to think that, that’s acceptable and the margin for what’s acceptable for the organization now has shifted. It’s almost impossible to recover that.
Daniel Serfaty: This is profound and this is true. And we are witnessing things like that on a daily basis. Those of us who study leadership because we like it and we observe our society. So I’m going to ask us to imagine for a second now the future. And in this particular podcast, in MINDWORKS, we ask a lot to that question about the degree to which AI is changing our world, making our world more complicated or simpler, alleviating our workload, but now I want to orient it towards leadership and leadership development. Are there techniques there that are really going to change or revolutionize even the way we develop our leaders. And then if we look at the practice of leadership to which maybe some of this AI is freeing up the leaders to do things that leaders do.
Morgan Darwin: Daniel, I think it maybe also useful to also add automation into this. So what I think that, this does create an opportunity for work to become much more interesting to many, many people, as AI, machine learning and automation began to take over the mundane aspects of work. And I think that, as we move into the future, leadership will actually become more important because the franchise of the human mind will become of greater and greater value rather than just the ability of human beings to toil. So I think from that perspective, leadership will become increasingly important if you want to add true value to the organization.
And so what we will begin to focus on is the development of the individual so that they can better create or better add to the success of the organization. And I really hold very strongly to the idea that technology of any type is there to enhance a human performance, not to replace humans. And I’ll be honest with you, I think that’s one of the advantages that the United States has going forward in the future in comparison to other nations, is the value of the individual of the human being in the process, in recognizing that the human being is contributing more than just a logical function.
Fred Diedrich: Yeah. I mean, Daniel, it’s a very hard question, but if I think about much of what we’ve talked about here, today we’ve danced around leader development, not only what makes a good leader, but what will make a good leader in the future. And so, to me one of the fundamental issues of this future that we might envision is going to be asking ourselves to question what are the AIs learning from us. And thinking about being a leader, not only of the people in that organization but the AI in that organization is pretty interesting. And we’re going to have ask yourselves what it is you want to be learned because we may or may not like what it is that gets learned by these entities.
And so in many ways, I personally don’t see it as being that different. I think it’s going to be different in things that people do, things that machines do. But at the end of the day, I believe that as we worry less and less about simple machines, less and less even about automation and more and more about AI, the very lessons that we’ve heard articulated by people like Morgan and Scott today will be the ones that will matter most when in fact we try to lead AI.
Daniel Serfaty: Thank you. Fred?
Scott Flanagan: We’ve had some of those conversations in the past from what’s AI learning from us? It is a difficult question. I guess my mind has been going to, it’s not so much a fear or a concern, in my little bit of experience, sort of machine learning and algorithms and things, I can tell you truthfully, that some of the thoughtfulness and some of the things I’ve done, I’ve sort of outsourced it to machine learning. I’m not suggesting that will completely happen with leadership, but I can see where maybe some of it would be outsourced to AI and that could be problematic. I brought up earlier in the conversation about what’s under the hood is most important. And so that’s where I see where leadership lies.
Morgan Darwin: It’ll be interesting to see whether AI, if it does begin to assume some form of leadership position, is able to assume the, I’m going to say the personality traits that we’ve been talking about today. Because in the end, we don’t know, in the end as a team satisfied with being valued by artificial intelligence, the same as it is by a human leader, can you establish the same level of trust? So I think that if AI does play a role, then the role of AI to be successful in a human relationship has to be one of trust, accountability, judgment, and so forth. And so I think that to a degree that’s a measure of reassurance, that it should be going the right direction, because if AI wants to have a relationship with humans, it has to be a good leader. And if it’s a good leader, then it’s accountable and values the people that it’s working with.
Daniel Serfaty: That is pretty profound. And I think that it’s linked to the question that Fred asked earlier, which is what is the AI learning from us. Because AI needs data in order to behave. And maybe-
Morgan Darwin: We have to set a good example.
Daniel Serfaty: Exactly. There will be some emergent properties in that AI that’s going to surprise us and maybe hold a pretty tough mirror to us as leaders to say, “This is what the AI concluded about or learned about your leadership.” That can be interesting, but certainly a new territory for all of us to insert those new forms of intelligence into our lives, especially with us as leaders.
Morgan Darwin: If it goes the other way, Daniel, we’ve got the huge adventure in front of us as we go into the matrix and have a choice between the red pill or the blue pill.
Daniel Serfaty: This is why Hollywood is always a refuge, of all the big questions for us. But Scott Flanagan, Fred Dietrich and Morgan Darwin, thank you very much for participating in this illuminating discussion on leadership. Having work with you directly and indirectly over the years, I know that you have exhibited great aspects of leadership, not just in what you learn and what you teach, but also what you do. So for that, I am thankful.
Thank you for listening to MINDWORKS. This is Daniel Serfaty. Please join me again for the next episode. We welcome your comments and feedback, as well as your suggestions for future topics and guests. We love to hear from you. You can tweet us at MINDWORKS podcast or email us @email@example.com MINDWORKS is a production of Aptima Incorporated. My executive producer is Ms. Debra McNeely. And my audio editor is Ms. Lindsay Howland. To learn more and to find links mentioned in this episode, please visit aptima.com/mindworks. Thank you.
Daniel Serfaty: Welcome to MINDWORKS. This is your host, Daniel Serfaty. Artificial intelligence is making its way into more environments, workplaces, and missions in our lives. But if a key goal of AI is to augment and improve how we, humans, perform our jobs, how are we preparing for this new reality? Specifically, what training will be needed for both sides, humans and artificial intelligences, to ensure that these new hybrid human AI systems and human AI teams could work optimally together. I had a chance to explore this fundamental question with three experts at the end of last year in a panel discussion at a conference called [inaudible 00:00:50], a very large training simulation and education conference. And like all great conversation it left me wanting to know more. So we are taking a break this week from our Mindwork series on leadership, because the opportunity arose to continue that discussion with my three very special guests.
Dr. Mike van Lent is the CEO of SoarTech, which develops human centered AI solutions for the military toughest problems. Mike has been very passionate about AI since he purchased his first computer more than 35 years ago, telling his parents that he was going to make the computer think, imagine that. Mike has published widely in academic journals and conferences and is the founding organizer of the AAAI Conference on Artificial Intelligence and Interactive Digital Entertainment. My second guess is Dr. Greg Zacharias. Dr. Zacharias serves currently as director of federal solutions at Pasteur labs, a company that was founded to develop and apply simulation intelligence in today’s large scale complex problems. He recently retired from government service as a chief scientist for the department of defense director of operational test and evaluation. He also served as chief scientist of the US Air Force creating a long term roadmap for developing AI enabled air force systems. He entered government service after co-founding and leading for many years, Charles River Analytics, an award-winning R&D small business focused on integrating AI with human systems engineering.
And third is Dr. Fred Diedrich, who is making a return appearance on MINDWORKS. He’s currently an independent senior consultant with more than 20 years of experience, improving human performance in a range of government sponsors, including the Army Research Institute, the Defense Advanced Research Project Agency, the Office of Naval Research and more. He was previously the CEO of Milcord as well as president and my dear colleague at Aptima Incorporated. Welcome Mike, Greg and Fred. So before we take a deeper dive into this nature of human AI collaboration, what happen if we don’t take into account the human side of the equation? In a sense that many folks in our audience say, I saw the podcast was about AI and here is a computer science and aerospace engineer and the psychologist coming in and telling me, we need to take into account that human part of the system. What happen if we don’t? What happen if we ignore that part? And as we are flooded right now with all kind of claims about AI, we let AI be AI. What are the dangers of doing that, or should we do that?
Mike van Lent: I’ll start here. One big danger that I think about is the possibility that we spend a lot of money, and time, and smart people’s energy building AI systems that never get used, even if they’re actually useful systems. And I think this is one of the key distinctions that I think a lot about. Just because an AI system is useful, it could help someone do a job, that doesn’t mean that person will use that system. I know I have built systems, both AI systems and training systems, that were proven to be effective, but for a variety of reasons, some cosmetics, some deeper, they didn’t pass the human sniff test. And so they didn’t go into use.
And so there’s a lot of potential to build AI systems that could be very useful. That potential won’t be realized unless we’re also thinking about how to make sure these systems meet the acceptance criteria of the users, and they go from being useful systems to becoming used systems. And so that distinction between a useful system and a used system, and the fact that becoming a used system puts additional requirements on the underlying AI, I think is really important and something that should get more attention.
Daniel Serfaty: Greg, Fred, you want to add to that question about doomsday that will happen if we let AI be AI and we don’t take into account [inaudible 00:05:13]?
Fred Diedrich: Yeah. Well, there’s many ways to get to doomsday, but I think it’s a great point about being useful and usable. And I think I might only add to that, implicit in that is that it’s hard to imagine that in military operations, for instance, humans won’t be involved. So in other words, the AI will be part of a team explicitly or implicitly. And so the effects of that team are going to be what matters. And so it’s hard for me to even begin to think about why we wouldn’t consider the AI in that larger context.
Greg Zacharias: If I could give you a episode from my past to illustrate what happens when you don’t consider this. So as a young engineer I got to work on the space shuttle, working on an autopilot that went from Mach 28 down to Mach 5, hypersonics, huge range in altitudes and so forth. And I designed this incredible wonderful system that followed the guidance commands and did all these wonderful things. I saved thousands of pounds of extra fuel and so forth, got it in the simulator and got the astronauts to test fly it. And every single one of them hated it. And they said, “This doesn’t fly like an airplane.” And I kept telling them, “Well, it’s not an airplane. It’s a blunt body doing a reentry.”
Anyway, it ended up being a failure in terms of acceptance by the pilot who was interacting with it, the set of pilots that would interact with it. It was not a case of AI, but it was a case of engineering and just did not fit the case of what they want to do. I might add, as a postscript, I talked to a bunch of astronauts that actually flew the reentry system that wasn’t mine and they never touched the stick actually. They just let the thing go automatically. So mine wouldn’t survived the sniff testing, but I learned from that way back when that you just can’t take the human out of the loop in these cases.
Daniel Serfaty: Yes. Thank you. So there is a human in the loop, on the loop. There is a lot of in the literature right now to understand in those complex systems where the human should be. Especially when we have, and you make the distinction Greg about even if it’s not AI, those complex system that have human in them need to model the human and take the human into account to be not only accepted, but also eventually maybe to behave optimally. But there is a debate and I don’t know if it’s a semantic debate, it’s a real debate that has implication about how do we approach the design of an AI that augments the human work, does it replace it?
We’ll talk about that notion a little later in the podcast. But should we approach the design of the system of the human system that is augmented by a capability that is AI based as a tool or as a teammate? Is that tied to semantic discussion or is that a real discussion that has implication for design? Who wants to take that, the teammate versus tool aspects specifically about AI? Because AI at the capacity of learning and adapting and changing, and it’s very natural for us humans to think of it as a teammate, like a human teammate will adapt and change and evolve with us. Want to take that on?
Mike van Lent: My first thought there is we, the designers of the system, don’t always get to pick. The users are going to relate to what we build in ways that are dependent on how we build it and may not be what we anticipated. And so humans have millions, tens of millions of years of experience working with other humans. And there’s a lot of deep social, but also biological structures in the brain around collaborating with other humans. More recently, humans started working with tools and whether something is in that human category or that tool category, is a judgment humans make all the time. And once we start building these AI systems that can speak, that can be making decisions, that can be understanding what the human’s trying to do in helping, it slips closer and closer to that another human or another intelligence and that evokes and activates a whole set of structures in humans about how to work with something.
And so some of the promise of AI is that AI can be doing these things interacting more naturally making decisions on its own under human’s direction. Once you start to take advantage of that promise, you inevitably slip out of the tool category and into the, I’ll call it, another intelligence category. And then that brings with it a whole set of expectations and assumptions and biases that you need to take into account. You get into things like the uncanny valley. Whether it’s the physical manifestation uncanny valley of how a figure looks, or the kind of decision making uncanny valley of, this is acting kind of like a person, but not really like a person. So I think it’s inevitable as we build towards what AI can do, that we’re going to start to be in this gray area between teammate and tool. And that means we need to take the teammate side of that seriously.
Daniel Serfaty: I’ll pick up on something you said and ask a question of Greg, in the example that you gave previously about your design for the shuttle reentry system, would the kind of design decision you would’ve made, if the control system, et cetera, was actually an artificial intelligence that had the capacity to adapt basically to the astronaut or to the user. Is there just a quantitative difference here in term of the computational ability of one system versus another, or there’s really a qualitative system in the way that Mike described. It was not just a tool for enabling the reentry, it was also some kind of collaborative teammates that will be given to the astronaut. Would that have been different?
Greg Zacharias: Yeah, I think quite different. I think the problem would be you would like the autopilot or the AI to recognize that the pilot was frustrated with the response of the aircraft. And that maybe a pilot or a co-pilot, maybe humans might have done that. And then the second part would’ve been much easier to actually adapt the control laws, which is what North American Rockwell did. They changed the control laws so that the pilot would make it feel it was an airplane. It cost an extra 60 million to launch in fuel costs, but that was a trade off they made. To do that in real time would’ve been maybe doable I guess these days, not back then, obviously. But that would be a good teammate’s response to help the person flying the aircraft I agree. So that’s a meta model development by the AI of your teammates.
Daniel Serfaty: Talking about that. Thank you for that last point. And that’s really a question for you Fred. As the developmental psychology of the AI develop, and to do that you’re going to try to embed or give the AI some kind of an internal model of representation of the humanities designed to support. Is that possible? Is that something you think about as a person that cares about skill acquisition and other things like that?
Fred Diedrich: It’s a good question. And in many ways gets to the heart of the matter. It’s certainly the case that in order to do the kind of intelligent activity that Greg was just outlining, you will need to have within the AI some model of the human. And in many ways, this falls under the umbrella of theory of mind, that is that the AI needs to have a theory of mind of its humans, and the humans I think need to have a theory of mind of their AI. So that goes both ways. And that kind of understanding will enable hopefully the kinds of performance gains that Greg was just alluding to.
One important thing to think about though, is that if we are assuming that the AI evolves over time, and it doesn’t have to be the case but let’s make that assumption in this case, and let’s assume that the humans evolve over time, then that theory of mind is not static. And so the idea here is that you can’t just program something in that we know that Daniel’s going to behave in the following way because Daniel might change. And I think that makes this kind of a hard problem.
Daniel Serfaty: I want to go back to something you said Mike earlier about this notion about supporting decisions and knowing where the decision is being made. Should the locus of decision be in a system that has, let’s say, one human, one AI or multiple humans, multiple AI. Should the locus of decision be always with the human side, or are you ready to accept as a scientist, the fact that sometime the locus of decision should shift to the AI side and the human should remain kind of an observer of that decision being made?
Mike van Lent: With respect, I’m going to say that’s the wrong way to ask the question. In my mind, there’s a distinction between a decision and a goal. So I believe the locus of the goal that the human AI team are pursuing should remain with the human. The human should set the goal and then it’s the job of the AI and human together to make decisions to pursue that goal. The AI, I believe, will be inevitably making decisions without consulting with the human, whether it’s because it’s a domain where there isn’t time to do so. You get into some elements of a cyber domain and the human just can’t be making decisions fast enough, so the AI has to be making decisions without consulting the human. Maybe there are communication constraints, other reasons why the AI can’t consult the human. What’s important is that the AI and the human be working towards the same goal.
So I would say when you’re thinking about who’s setting the goal, the locus of defining the goal needs to rest with the human. The decisions to pursue that goal, some of those the AI can be making on its own, some of those the AI can be consulting with the human. But what’s really important is making sure that both sides of this team are working towards the same goal. And that gets a little bit into what we were just talking about with theory of mind. If the AI has some goal that is learned by the system, that’s not represented in a way that a human can inspect, then you got some questions. You don’t really know what goal the AI’s working towards, if it can’t say here’s a human understandable representation of the goal I think we’re working towards. When I talk about the underlying algorithmic constraints that you need to build an AI that is usable by humans, that explicit representation of a goal that a human can inspect and confirm is right is one of the kinds of things I think about.
Daniel Serfaty: Good. You anticipated my other question, which would’ve been the right question this time, which is how do I know that while working with an AI, that the AI is actually listening and understand my goal? All I can observe is basically the outcome of its decisions or its insight. Not necessarily an acknowledgement that whatever its recommendation or its decisions are, are compatible with my goal.
Mike van Lent: I think different AI systems differ on that. Some the goal is implicit deep in the system, others the goal is explicit and you can communicate it in English. And there have been lots of research, even recently, about how do you take the implicit goals and try to extract them from the system and create an explicit representation of them. Some of those involve testing the systems, some of those involve other approach. I think that’s one of the core challenges right now in the AI research behind the field is, how do you understand and verify, validate what the goal that these systems are working towards? Even if it’s explicitly stated, do you have confidence that this AI is really always working towards that explicitly stated goal?
Greg Zacharias: I think this is a little more complex and I don’t want to take a couple points with my point. I think we’re thinking of a team of almost equals. I’m giving you a counter example where you’ve got the Air Force team of 700,000 people and they’re hierarchized. And it bring to mind Daniel’s earlier comment about a person in the loop and a person on the loop. A few years ago I proposed we have a person under the loop.
And for instance, you may want to have an AI that runs the air operation center, telling the pilots what to do. And there’s so many levels down in the hierarchy. They’re under the loop basically of the AI, but they’re all part of the same team. So I think that’s a dimension that we got to bring into this equation. And I think the other thing Mike already mentioned was speed of response. You want the auto G cast ground collision avoidance system to take you out of a bad situation, you want it to work in millisecond’s time when you’re not even paying attention to where the train is or in cyber as well. I hate to say, I think it depends situation where the locus is.
Fred Diedrich: One of the things that the various comments make me think about is that we’re dancing around the ideas of constraints. So one of the ways to think about this is to provide some constraint on the way that the system might evolve that lines up with your goals or not. And what I would point out is that we already do this in the following sense. So if you think about commander’s intent or an order that specifies purpose key tasks and end state, those things are set up that way intentionally. Because that brigade commander, for instance, is giving his intent, he’s not going to necessarily tell the company commander exactly how to do his job. But what he’s doing when he gives intent, purpose, and end state, is he’s laying out the constraint space while giving flexibility to the entity to realize that.
And I suspect some of these ideas that have been pretty well laid out for human teams, may in fact be informative here, not withstanding the really good question and point about how do we communicate that. So there may be communication that could even be verbal, there may be some ways to measure certain aspects of actions that tell us something about whether or not we’re within that constraint space or not. And that might depend, as Greg said, on the context.
Daniel Serfaty: I think what fascinates me in the subtext of this decision, we don’t have all the words yet. We don’t have all the vocabulary necessary to understand this emerging interactions. Some will say with different intelligences, some other people will even dare say with the new species in our midst. And in a sense you bring the analogy of management by objectives, what people, at least in the business schools are studying, they don’t use the same terms that is in the military, but that notion that you don’t micromanage, you just set goals. And we each try to use different analogy either from control engineering or from team theory in order to be able to explain that relationship. And I think it’s fascinating. I think that the English language is not rich enough now to have all the words necessary. Even the word intelligence, maybe somebody will come up with a different word to explain what exactly are those systems that we are building. They’re certainly not human intelligence.
Let’s go down a notch for a second. I’m going to ask you to provide for our audience any examples you’re aware of. Either because you participated in, or you witnessed it, or you investigated it, the case in which there was a successful, or there is a successful example in which AI and humans have learned to work together, having a successful design basically of this human AI system. I also would love to hear about unsuccessful one because that’s the one we learned the most of. Where automation and humans, or intelligent automation and humans, didn’t cooperate the way we expected them to cooperate, maybe leading to an accident to a disaster. So both good and bad examples for our audience. Who wants to share them?
Fred Diedrich: Greg, do you want to kick us off this time?
Greg Zacharias: Sure, I’ll try. Well, I think the autonomous automobiles are getting to be a successful example. I think they started out with everybody keeping their hands off the wheel and sitting back and still in instances of it. But I think people are learning on how to interact with them. And I think that’s going to be a success over time as people learn their limitations. And it goes back to the locus of control. I think most people have control over the mode that automobile is in rather than the other way around.
I do think example of unsuccessful thing. Again it’s not so much AI, but I think AI will be subject to the same kind of poor design. The case of the Boeing 737 MCAS augmentation system, which was cases of expanding this flight envelope where it shouldn’t be. Having a flight critical system depending on a single sensor. And then, as a friend can dive into, not providing adequate training to the pilots that we’re supposed to be flying that. In fact, hiding the existence of the system from them, which is a really egregious kind of case. And certainly it could have been an AI system doing all this too. So I think that’s just bad engineering design. Over to you Fred.
Fred Diedrich: Thanks. Those are great examples. And you might say that in the case of the aircraft, the pilots, you were not trained, didn’t have a good theory of mind of the automation, which speaks back to something we talked about a moment ago. So it’s a great point. Color commentary and autonomous cars. I think based on what I observe in Boston on a regular basis, the bar is very low. And so what we might call success can be easily achieved, potentially, hard engineering not withstanding. I think that there’s so many great examples. And I’ll throw out a really simple one I think everybody can relate to.
Every weekend on Saturday morning I hop in my car and I put my phone into the car and the car pops up a map that’s taking me to the supermarket. And so my iPhone here is learned that I go to the supermarket on Saturday morning and it’s mostly right. Sometimes I’m not going to the supermarket, so sometimes it’s wrong, but there you go. I don’t know if it really helps me or not, because I kind of know how to get to the supermarket. That’s maybe another question about how useful it is, but it works in a sense. Another example that I’ll throw out that’s received a more popular attention in the media was the idea of AI and candidate screening or mortgage lending practices and whether or not the AI might reflect inherent biases that it learns. I think that’s a particularly interesting point because it raises questions about whether or not… Well, let me put it this way, what the AI might be telling us about us, which raises some interesting questions, but there are a couple examples. So Mike, anything else you want to chuck in?
Mike van Lent: Yeah. So when I think about the big successful AI systems, I think about things like the Google search engine, spam filtering and detection in email systems, recommendation systems for things like what movie you should watch given the movies, or what you should buy on a website given the things you bought. What’s interesting about all those is, it’s lots and lots of data, a lot of it mined from other human decisions and it’s relatively low stakes kind of decisions. It’s not the end of the world if a spam email slips through the filter or I watch the first 10 minutes of a movie that I don’t like and then turn it off. You look at the other side which have some of the same properties, and one I’ve seen recently is the example of Zillow. Zillow recently invested a lot of money in buying properties based on its AI algorithms estimation of how much those properties would be worth going on the market.
That was a much higher stakes kind of example and didn’t go well for Zillow. And part of that is I think they were making predictions about an uncertain future that’s a little different than making predictions about a more certain whether this is a spam email or not. So I think there’s lots of ways where all of us are using AI every day and it’s working very effectively. This dips a little bit into one of the jokes within the AI community is that once it starts working and people see it working and understand how it’s working, it’s no longer AI. Then it’s just a search algorithm or it’s just a spam filter. And AI is always the thing we haven’t done yet. So there are a lot of really positive examples of AI out there, but usually on lower stakes kind of decisions at the moment.
Daniel Serfaty: I love that comment. I’m going to remember that and reuse that with a proper credit Mike. Thank you. So because some of those systems that you just described are out there, like the Zillow example, they range from something in which there is no really human intervention in a sense. It’s an algorithm that goes out there makes a decision based on future uncertainty of the market and either succeed or doesn’t succeed in term of some key performance indices that people calculate. As opposed to one that is much more personalized, much more individualized, like car driving there are some general behaviors that you would expect. And car driving when Fred steps into his car, his future semi-autonomous car, and the car will have already an eternal model of theory of mind of Fred and will behave accordingly. Either by speeding a certain way, driving a certain way, sharing data a certain way, et cetera.
On that last category. I wonder if this notion of working with AI is a new competency in the 21st century. It’s not working with a hammer or an Excel spreadsheet, but in and by itself, working with a system that adapts to you, that personalize its own functioning to you, whether or not it’s a new skill. It’s a new skill that we have to put much more effort in developing either in the general population or in our schools. If indeed that report that I mentioned in the introduction is correct, more than 800 million jobs are going to be displaced. It doesn’t mean they’re going to be eliminated, but they’re going to change as a result of the introduction of AI, at least change if not more. Are those new skills and if these are skills, do we know how to train them? I’ll start with you Fred, because this is really your domain but I would love to hear also from Mike and Greg about that.
Fred Diedrich: Yeah. Thank you. I don’t think it’s that new, but it may be different. And I think early on in the conversation Mike used the term other intelligences. And so if we think broadly already today, we all work with other intelligences. And the examples we just went through kind of are on some kind of a continuum from pretty sophisticated to not very sophisticated. And I think that the key from the human perspective is that we must be “adaptable or agile.” And a lot of our sponsors talk about this and have talked about it for years. And so I’m not sure if it’s a new competency, but it is something that I think will stress the competencies of critical thinking, problem solving, et cetera. And we need to probably ask whether or not, many of our traditional educational programs, they’re in the military, business, or academics, really do indeed rely on and teach those kinds of critical thinking skills that we think they do and they may not. So maybe it’s more of doing what we’re supposed to be doing already is maybe the way I’d answer that.
Daniel Serfaty: Greg and Mike, you want to chime on this one, are those new scales for humans?
Greg Zacharias: I’m going to be a bit of a curmudgeon here and think of, again, going back to this theory of mind. I think we are such a social species. I think we have this theory of mind built in over millions of years that we expected another human to behave, kind of like we do. And when we’ll be set with pathological people who are sociopaths, we don’t even know how to deal with them. And I think for these systems to be accepted, they’re going to have to become more like us rather than us figuring out them, hard wiring to deal with it. Now that’s not my curmudgeonly part, what my curmudgeonly part has to do with DOD’s approach to this. And I have a quote here. It’s one of their ethical principles that was kind of designed by the Defense Innovation Board and signed off by whichever sec [inaudible 00:30:04] we had at the time. The ethical principle for AI and call for traceability.
I’m going to quote this, “The department’s AI capabilities will be developed and deployed such that relevant personnel,” that’s the key phrase here, “possess an appropriate understanding of the technology, development processes, and operational methods applicable to AI capabilities. Including with transparent and auditable methodologies, data sources, and design procedure and documentation.” So here I’ve got a pilot or an army guy firing a smart missile, and they’re supposed to know all this stuff. And given that PhD researchers of deep mind still don’t know where some of those go moves came from, we’re going to expect some PFC to understand what data set was used to train this and what this thing. So I don’t think that’s going to go anywhere. And I think really it’s going to go the other direction. The demand is going to be, I need systems to [inaudible 00:31:00], behave like humans.
Daniel Serfaty: That’s an interesting insight about that. Mike, do you agree, are we designing AI to our image?
Mike van Lent: I think we are. I think and Greg said this. The key phrase in what he just read was relevant personnel. And I think how much that pilot or that infantry person at the end of a long chain of system development, what they’re the relevant personnel for versus all the others in that chain is going to be very interesting. You can’t expect a pilot, for example, to know every aspect of the system. And if you’re building the system that requires that you’re building the system wrong, I would say. You can leverage this stuff we all already know about how to work with other humans and take advantage of that.
Ideally, and I think this perfect world is not going to exist, but ideally you could leverage a lot of stuff people already know about how to work with others and not have to require them to learn a whole new skillset. But that ideal case is I don’t think ever going to be a hundred percent achievable and the relevant personnel are going to have to learn some new things. The question is can you keep that to a manageable level, given everything else they already need to know to do the job they’re relevant for?
Fred Diedrich: Yeah. It is an interesting point about moving towards something that seems more human like, or is human understandable as a way to kind of deal with this. And I’d point out that within the currently existing leader requirements model for the army, one of the attributes we want people to have is interpersonal tact. And the reason why that’s in there is in part, because you have to know how to adapt to other kinds of folks. So that’s not a new attribute, it’s an existing attribute in this case that interpersonal tact might be with respect to some kind of artificial intelligence. But nevertheless, it’s the same problem. And I’d add that what makes interpersonal tact hard is to have that model of the other individual that guides your adaptation towards it, which is really the problem we’re talking about here.
Daniel Serfaty: Yes. I think that to add to that I can share a personal experience. One of the very first edition of this podcast was actually with an artificial intelligence called Charlie that was developed by one of my teams. And in the process of learning to interview Charlie, I did a lot of this, how do you call that, interpersonal tact? Empathy, in the true sense of the word, empathy, to try to understand the others. It’s the same, but it’s also fundamentally different. The same way my ability to interview Jennifer, or Mary, or John is different because I have a different mental model in the sense of each one of these individuals then I had Charlie, which is an AI. But my adaptation was different the way I adapted to Charlie to ask her, that’s a she in that case, a question so that she gave a decent answer and not some stupid answer. It’s different because you could sense that the stuff that was coming back was not human. Sometime it was brilliant and insightful even beyond what the human would’ve answer. And sometime it was pretty pedestrian.
And that kind of variability is not expected from most humans. And I know you’re going to make a joke that some people are like that, but no, we are pretty consistent in a sense. And therefore that interpersonal tact, I think, goes into another level of difficulty, let’s put it this way, when you deal with artificial intelligence. So far we’ve explored a question about what we called about learning to learn together, and that together has to do with this notion of teaming in a sense with a technology that is intelligent, that is adaptive, that is learning. I want to explore whether or not there is here an opportunity for true synergy or is just in a sense, a marriage of convenience. It’s just something that is out there and we better deal with it because it’s inevitable. Or there is something very optimistic about those new possibilities about transforming the very nature of human work, whether it’s in the military or in business in the workplace.
So my question for each one of you is where do you believe are the low hanging fruit? Where do you believe this notion of really understanding, and designing, and testing, and experimenting with human AI teaming or human augmentation with AI, is the most promising? Is it in defense, an area that all of us know quite well? Is that in healthcare? Is it in transportation? Perhaps even in education? Can you just speculate about it. Where you have to predict the next 10 years, where is that going to make a big difference? Pick one domain and expand on.
Fred Diedrich: Well, I can jump in and stealing something from Mike, but I think he pointed out something I think pretty interesting earlier. I think it was Mike, maybe it was Greg, but there’s a difference between high consequence environments and low consequence environments. And so the example I believe that was used earlier is if Netflix recommends a movie to you and it stinks, so what? So the barrier to entry is very, very low. And I think that’s an interesting way to think about it. In terms of where you might see the biggest impact, maybe in some areas where even already it’s really not that consequential. Now, as we push into the kinds of domains that some of us work in with the defense and stuff like that, it could be very consequential and maybe call for us to be more conservative. The problem with that’s going to be that “the enemy will have a vote.” And so the timelines of these things may not be fully in our control. In the sense that you may be surprised about where some of the different domains really make a lot of headway out of necessity.
Daniel Serfaty: I see. And when you say the enemy has a vote, can you expand on that a little bit, what you meant by that?
Fred Diedrich: So if we think about speed of operations being very important, and this is an example that Greg talked about earlier, I would start there. Given the amount of information coming at us, given the speed at which the enemy would be moving, we will be put in a position where we have to be faster and faster, or at least find a way to contend with that. And so that’s what I mean by, in this case, the enemy will have a vote because some of these things may cause us to react in different ways.
Daniel Serfaty: Understand. Greg, Mike, you want to venture a prediction on the domain that you either like or you’re familiar with?
Greg Zacharias: I’m happy to. So I think the consequences, the size of the consequences, is clearly going to be a driver. I think the size of the organization is another driver. I really struggle with what DOD is going to be doing in terms of accepting and bringing in AI. I listened to Eric Schmidt at AFA a couple days ago. He raked a whole bunch of four stars that we’re sitting in front of them saying that, “It’s clear that you can’t get your bureaucracy to bring in AI directly. So what you’ve got to do is have a whole bunch of little pockets of 50 people distributed among your 3 million people to do that.” And that’s happening now, but how that actually gets embedded into larger scale systems is a big question. So I’m saying these are probably not where things are going to happen. I think that they’re going to be happening in more individualized distributed places.
Then you think about autonomous vehicles. There’s one person in their car, there was really no attempt. They’ve set a network of such cars around even though now you have Tesla vehicle A telling vehicle B that there’s a pothole somewhere. That networking of this is happening, but that certainly wasn’t the darker challenge at all. It’s just a much more simple 1v1 or one with one team. So think those kinds of applications are more likely to happen than for large bureaucracy, even though in an air operation center would be hugely improved if we apply some of these approaches to that.
Daniel Serfaty: So the grave consequences actually is not an acceleration proposition for implementing AI. The bigger the consequences, the more time it’ll take for AI to be thoughtfully integrated.
Greg Zacharias: You’re more likely to see it in an individual aircraft than in stack headquarters in office.
Mike van Lent: I certainly agree that the consequences is going to be a big driver. Another driver I think about is, what is the alternative to AI? So you look at something, a domain like transportation. The alternative to AI is a human driver. There are millions of human drivers out there in the world and the cost of a human driver isn’t super high. You can look at Uber and say that Uber is just an automated AI company where they have a human in the loop doing the job of the eventual future AI. Uber business model maybe works because they can get those human drivers pretty cost effectively. And so in transportation, you need to build an AI that’s more cost effective than a human driver, and that’s going to be a challenge. Another area I look at is the availability of knowledge.
AI systems they need knowledge. Whether it’s a big labeled training set they can learn from, or a human who can teach. them how to do something. The knowledge has to be there and I think that to me is a real challenge in the defense domain is, we have a lot of doctrine and stuff but it’s very expensive to go out and conduct these large military operations to gather the knowledge. And a lot of this is such a dynamic domain with so much on the flight decision making that I worry about where all that knowledge is going to come from. So then I look at healthcare. Healthcare the alternative is a trained medical professional, which are very expensive to create and in low supply. So the alternative to AI is hard to get. There’s a lot of knowledge out there. That’s an area that I think is ripe.
The challenge of course, is those are very high consequence decisions. And so the availability of knowledge and the in availability of a good alternative flies in the face of the fact that it’s high consequence decisions. But I think certain areas of healthcare, especially if it’s a decision that a human can then check is an area that I do think is primed for a big impact. Now, if you take impact in a slightly different direction and say, where is the introduction of AI going to have an impact on society? If AI takes over the transportation domain and tens of millions of people are out of the job as a result of that. Every truck driver, every taxi driver are out of the job, that’s going to be a huge social problem for the world created by AI. And that’s a whole nother area of problem, but I think that’s an impact potentially in the negative sense that AI could have. That might be very disruptive.
Daniel Serfaty: That last point, actually, thank you Mike, and Greg, and Fred, because you gave some good examples for the audience to think about their own domain. Whether they are teacher or a military war fighter to see, okay, how can AI change my life, change my work? The last sentence that you shared with us is actually where the number one question I’m sure when you guys are lecturing the same thing happened. But when I give presentations or lectures about this topic, that’s the first question people ask. Is AI going to take over? And they will kind of jokes to answer that question, but it is a serious question that’s created serious anxiety in some people.
Daniel Serfaty: Because only model they have is the Hollywood model, basically of the science fiction model of robots taking over, going wild, not behaving as expecting, et cetera. It’s not just a question of a replacement, it’s an unwanted replacement. And talking about that brings all kind of topics of issues about ethics. What is ethical AI? Not just in term of the single system, how a single system should behave fed by the knowledge that we fit it, as you mentioned, Mike, but also generally speaking for society. There are some major ethical questions about the design of AI, the introduction of AI, and frankly, the proliferation of AI. Can you comment on the one that we’ll review the most right now?
Mike van Lent: So for me, it’s that social disruption of the wide scale introduction of AI into a certain industry. There are going to be some industries where AI is going to do the job that a lot of people currently do, and that’s going to be a real challenge to me. That’s the thing that concerns me most in the near term next 10 or 20 years when I think about AI. We have the legends of John Henry, some guy who really good at hammering a spike into rock. And a machine came in and took his job away. That’s deep in our folklore. Who is the John Henry of the AI world?
Daniel Serfaty: So even at the societal level that is going to provide a disruption. Fred, Greg, any other ethical worries?
Greg Zacharias: I think it’s far too early to worry about ethics factor. Cases about biased data sets and so forth that we’re concerned about with judging people in terms of their credit or their ability to graduate from college, et cetera. I think those are easy things to solve, but honestly, we’re still trying to get these systems to work basically without any ethical constraints. But I feel we’re in the Wright brothers at Kitty Hawk stage when someone would tap them on the shoulder and say, so what do you think about incendiary bombing of Tokyo and what your technology is going to afford in the future? Maybe you should get out of this business. I think we’re just at the very beginning. And I’m concerned, frankly from the DOD standpoint, because there’s a lot of people working on this problem right now. Frankly it’s a policy [inaudible 00:45:19] in DOD that are trying to drive the train of the technologists. I think that’s an interesting situation for DOD, frankly. Over.
Daniel Serfaty: Listen, Greg. I don’t know if it’s just DOD. Your own Alma mater, I’m going to out you here as an MIT graduate. There is a relatively new school of computing at MIT, which is different from the school of engineering, has an entire department, actually and associate dean responsible for societal and ethical consequences of computing in general, let alone AI. And they’re teaching the new engineers and new people coming out of that school about that. They have to take those courses in ethics and societal implications. You’re saying the worry is way too academic at this point, let’s just build first and worry about that later?
Greg Zacharias: I’m an engineer. So I would build first, let’s see what the capabilities are, and then worry about how you control these systems. I just think we’re so early on. It’s true, engineers never were given those ethics courses. It’s just mostly in the business school. I don’t know what to say on this. I just think we’re tying our hands way too early.
Daniel Serfaty: I knew that if I bring MIT to the table, you’ll listen to me. All right, Fred.
Fred Diedrich: I’ll answer it slightly differently. Again, relying on maybe more of a future vision. So Greg makes an interesting point that right now the kinds of problems we’re facing may be different than the ones we will ultimately face. What I think about is that every AI before it gets its training data is effectively, depending on the kind of AI we’re talking about. So I should say it that way, depending on the kind AI we’re talking about, and AI starts as an infant. That is it has the ability to detect patterns and in a sense, I don’t know if that AI has ethics at that point in time. But what I worry about more is that we need to think about what it will learn as it starts to grow from being its infant AI into its adult AI. And we know that humans, many, don’t get all the way to where we might want them to be on an ethical scale.
And so I think that this is in some ways a developmental question. And what I really worry about more than anything is that we think about an AI this way, the AI is in some ways can be a mirror of humans. And when we see the AI as a mirror of us, we may not like some of what it is we see that AI learns. And so the way I’d answer your question is I worry about ethics and I think we have to worry about it now, but really it’s not just an engineering question it’s a learning question. And that puts it in the realm of what I would call developmental science.
Daniel Serfaty: I’m glad you brought that point because I wanted to ask the question from the perspective of the dynamic system. In a sense, you are looking at a loop that is over the lifetime almost of the AI. And this is where we start bifurcating from the traditional human machine interaction into human machine collaboration. When there is that intelligence that absorbs data observes the human it’s designed to support. And as a result, changes and evolve and perhaps develop a new level of competency, maybe even knowledge about the world. What can we do today?
Especially as we look into the future, perhaps not with today’s system that don’t have, perhaps that level of sophistication. One to guarantee that the patient doesn’t go into direction that are not under the control of us. On the other hand that we continue to co-evolve with the system, with AI system. I want to take your development or your infant development analogy a step further in the sense that in the future, we are going to have perhaps an AI sitting on our desk, or multiple AIs, augmenting our work and we’re going to grow with them. What are some of the issues there that we should keep in mind?
Fred Diedrich: Yeah. Interesting point. I’ll tell you what I think in a second, but let me ask you a question. So you have a couple of children and they are intelligent entities who are evolving, who you don’t fully have control of. So what do you do, what are you doing?
Daniel Serfaty: You love them, and you nurture them, and you trust them. But I’m using those term very carefully because first it’s true, but also what’s the equivalent again for that other type of intelligence?
Fred Diedrich: Well, I love the answer you gave. So you started with that you love them. And at the risk of getting too crazy here, sorry, we can think about what it might mean to “love your AI”. I’m taking that to an extreme, but let’s run with it for a moment because what I’m suggesting is that let’s say you want an AI to learn to become an entity that operates with ethics. Part of the way that will happen is through social learning. And it will be watching you, it’s human, in terms of how you behave. It’ll be learning ethics in a sense based on that relationship.
So the fact that you might treat your children with love and respect and empathy, even though within the AI world, we need to think about what love and empathy and respect is. I would argue that there will be correlates that will help us develop that AI in much the same way that we try to help our children develop. And the thing is that’s hard too. Sometimes development’s a little bit wonky. Things happen in funny ways, go in different unexpected directions, but you try to support and interject and control in different ways over time.
Daniel Serfaty: You made me think about a bunch of things right now, but I’m curious to see Greg’s and Mike’s reaction to this developmental methodology approach. Greg, you are the no kidding band metal engineer here.
Greg Zacharias: So I have a couple issues and I think Fred is right conceptually, but practically we’re not in the same space to build systems like that. We have AI systems that need a million instances to learn something. And we as humans, we learn with two instances, maybe as two year olds. So there’s a huge difference there. There’s no causality in these things, they’re not building causal models. These are perceptual models that are pattern recognition, demising over a cost function that someone told them, I want you to recognize cars or bicycles or whatever it is. And certainly not to be empathetic with your human teammate. And we don’t even know how to build a metric like that.
And then by the way, if I need to do a million instances, Daniel I’m sure you’d be really happy to sit down for a million instances and say, well that really hurt my feelings, or that doesn’t seem ethical to me or whatever. We don’t know how to do that. Honestly, I worry about this deep learning neural net paradigm that everybody is building on going into yet another expert system, like dead end that we saw in the 80s that really doesn’t get to intelligence or teammate or all those things. I think causality building up mental models of myself, of the world, of you is really fundamental to being a good teammate and addressing these other issues that Greg talking about. Over.
Daniel Serfaty: Thank you, Greg. Mike, you want to chime in on this issue?
Mike van Lent: I think Fred brings up a super interesting point that I’m going to take in a slightly different context. Most tools, you build the tool and then you use the tool. And there’s a clean point where you finish building the tool and from then on, you’re not continuing to build the tool, you’re using the tool. A child, you don’t build a child and at some point you say, okay, this kid is built. Now they’re going to go. You’re always, and they’re always, developing, learning, growing. And Daniel, exactly what you said, you love them, you mentor them, you develop them, and sometimes you let them make mistakes even if you know it’s going to be a mistake. And the reason you do that is your job as a parent is not to make sure your child always makes the best decision.
Your job as a parent is to create a positive, valuable person out in the world. And sometimes letting them make mistakes so they can learn, even though in the right now it might be the wrong thing, in the ultimate goal it’s the right thing. And so AI, especially AI that learns, raises an interesting question. Is this something we’re going to build and then at some point it’s going to stop learning, and after that we’re going to use it, or do we want AI systems that are continuously learning and growing like children do? And I think making the decision about that is going to be really important to determine how we should think about these systems.
Daniel Serfaty: I love the way you guys circle that very complex question. Because I think it’s pretty fundamental if we design system today to start even thinking about those topics, even there far in the future. My guess is that future is going to come at us real fast once some AI is going to misbehave, so to speak or behave outside the design parameter of the engineer will design them. And people are going to start asking those questions, so I’m glad we’re asking them first year on the podcast. I have one last questions for each one of you and it’s really a request for advice.
Our audience is probably wondering, especially the young people in our audience, the students and the folks that are still wondering about this field, how do I become like this gentleman? What can I do? I am starting my college right now. I am in graduate school, perhaps. What kind of disciplines? What kind of courses? What kind of classes can I do? What kind of choices even in my career can I do to go into that field of human AI systems, which is an emerging field or what some people call today, human centered AI. So we are asking you through you as the three wise man here to give some advice, to focus in the audience about what they can do to grow and eventually succeed in this domain. Dr. Diedrich, Dr. Zacharias, Dr. Van Lent?
Mike van Lent: I’ll start. So the short answer is learn to code. The slightly longer answer and more inclusive answer is learn to create via computers. Whether it’s learning one of the big programming languages or learning how to create HTML pages or high quality Excel spreadsheets, it doesn’t matter. The point is no matter what field you’re going to go into creating via a computer is going to be an asset. And I think we’re shifting from that being a specialized field to that becoming a part of every field. Every kid, when they’re growing up, they learn math, they learn how to write. And I think you need to add elements of creating on a computer to that.
And that’s important, not just because it’s a thing you need to know, but once you understand how what you’re seeing on a computer was created, it demystifies and opens you up to a whole new world about thinking about that technology. So I’ve had the experience of my children, they learn a little bit of HTML. They write their first webpage and then they go to another webpage and they’re like, wait a minute. This has created the exact same way? And I’m like, yeah. And they’re like, it’s not magic. So learning how to create via and on a computer, I think is just a really important thing that everyone should spend some time on.
Daniel Serfaty: Wonderful advice. Thank you, Mike. Greg?
Greg Zacharias: So I agree with Mike. That’s where the 90% of the creation is going to happen. It’s on a machine, that’s the implementation. I also think from my own background, I think a systems engineering view of the world where you draw a little block diagrams, this is connected to this and those kinds of that ability to dissect the world into sub components and then look at their interactions. Whether you come from a mechanical engineering EE arrow whatever, that’s helped me a whole bunch.
Because I’ve diagrammed human systems with those blocks and have gotten lots of feedback, both positive and negative. And then I think the third thing was, I think almost every single behavioral site course that MIT had to offer, which was probably three, and read a lot. I basically did an experimental site PhD. I didn’t know it was that at the time, but it was. Saw all the issues and problems and difficulty in doing that. And someone said, it looks like you’re going over to the dark side from engineering and the site, which I’ve loved ever since. So I think those three areas is what I kind of built my own life on.
Daniel Serfaty: Thank you. And representing the dark side, Fred.
Fred Diedrich: Speaking for the dark side. Very interesting points. I certainly agree that some of the fundamental skills that were just addressed by essential starting points. But I’m also reminded of, there was a very famous psychologist named Carl Rogers. He did a lot of influential writing stuff I’m thinking about was from the late 1960s. And he gave a lecture, I wrote a paper forget what it was, where he talked about his fear of what was happening with graduate programs and psychology at the time. What he said essentially was that he worried that well methodology and quantitative skills were very, very important. Psychology programs were making the mistake at the time for not deliberately selecting for and emphasizing in their program’s creativity.
So what I would add is that I agree that the fundamental skills are very essential, but I would also encourage folks to find an educational program, whatever it is, whatever discipline it is, that not only gives you those building blocks, which you have to have, but that puts you in an environment that actually facilitates creativity. And if you do that I think that some really pretty wonderful things can happen that in the end may be far less important than whatever discipline it is you choose to study, say, especially as an undergraduate.
Daniel Serfaty: Wonderful advice, thank you so much. I’m sure our audience will listen to it and apply it in their own lives. Thank you so much, Mike, Greg and Fred, for really a very interesting conversation that took some directions that were unexpected at some point. I’m sure that the audience will enjoy basically the breadth of experiences and even language that you use here to describe this field. Thank you for listening to MINDWORKS. This is Daniel Serfaty. Please join me again for the next episode. We welcome your comments and feedback as well as your suggestions for future topics and guests. We love to hear from you. You can tweet us at Mindworks podcast or email us at firstname.lastname@example.org. MINDWORKS is a production of Aptima incorporated. My executive producer is Ms. Deborah McNeely and my audio editor is Ms. Lindsay Howland. To learn more and to find links mentioned in this episode, please visit aptima.com/mindworks. Thank you.