of influence is only expected to increase over time.
I am emphasizing this omnipresence of algorithms today and in the future since many people in reality have no clear idea of what algorithms are and, as such, miss out on how we should assess the real value and application of algorithms in our work settings. This is an important observation to note because businesses seem happy to embrace the idea that leadership by algorithm will be the next evolutionary step to take. And if we plan to delegate the influence that leaders have over others to an algorithm, then we also need to understand who these leaders are.
If we think about leadership, we quickly arrive at specific types of individuals with specific skills, or, in other words, we have clear expectations and views on the identity of the people we consider to be leaders. Because we have those expectations, humans are able to quickly decide what kind of leader we need when situations change. Or, to put it differently, human psychology works in such a way that when a situation with specific demands presents itself, we can quickly infer the kind of leader that is needed. Given these situational demands, we will easily accept a leader with certain skills as the one in charge and comply with their directives.
Business today is confronted with much uncertainty, a volatile market, and rapid changes requiring leadership that is able to deal with complex and ever-changing situations. Algorithms and their unique capabilities of being rational and consistent in dealing with complex and highly ambiguous events seem to fit the bill to lead in such business situations. In fact, as I have documented earlier, today’s changing business environment is making the algorithm a prime candidate for tomorrow’s leader.
But is this really true? Can we make such a decision based on our perceived understanding of what the situation demands and the abilities an algorithm presents? Especially if we are not even clear how the algorithm is to be defined within the context of our society?
When it comes down to putting algorithms on the world stage as leaders-to-be, we need to become better informed about the real leadership potential that this new technology has before we blindly commit to automation. For effective leadership to emerge, one necessary condition is that any future leader is able to offer an identity that can be trusted. It is the presence of trust that makes people voluntarily engage in an open and co-operative relationship with the leader. It is only when relationships are characterized by open-mindedness and collaborative behavior that the influence of an effective leader can kick in. Put simply, if the actor placed in the leader role is not trusted, no leadership can emerge.
Being wise, not smart
So, how do algorithms fare in this respect? The late Peter Drucker once noted that “the computer makes no decisions; it only carries out orders. It’s a total moron, and therein lies its strength. It forces us to think, to set the criteria. The stupider the tool, the brighter the master has to be – and this is the dumbest tool we have ever had.”53
If Peter Drucker’s wisdom proves still to be true today, then we need to worry, because a potential leadership disaster may be circling above our heads. If algorithms move into a leadership role and fulfil the tasks dealing with a rapidly changing world, then a problem is likely to occur. That problem will center around whether or not algorithms do possess the skills to acquire influence to lead others. After all, leaders today need to be influential as they have to develop truly global organizations that operate effortlessly across borders. To achieve such influence, scholars have argued that a sense of wise leadership is needed.54 Are algorithms capable of doing so? If they are not, then we may have made algorithms appear to be wiser than they really are. And if this is the case, we need to be more careful in our assessment of how and when algorithms can be used in matters of authority.
Building on this logic, an important question to address is whether algorithms can really be wise while not being human? Again, research can help here. What we know so far, is that studies have shown that people perceive machines in general, and algorithms more specifically, as non-human. We perceive them this way for the simple reason that we are unable to attribute a “complete mind” to a machine.55 We do not consider machines and algorithms to possess the fully-fledged emotional (experiential) and thinking (agency) capabilities of humans. You may ask yourself whether it is necessary that machines need to possess the entire range of human emotions and cognitions for them to assume the role of leaders. Well, understanding that people only follow leaders if they perceive them to be legitimate, and that legitimacy is inferred from our perceptions of whether someone is wise, fair and mindful, then it is indeed necessary.
Research in psychology shows that – as humans – we only consider someone or something to have a mind when we can attribute both agency and experience to them.56 As algorithms are perceived to be limited in their abilities to show empathy or even understand the true meaning of human emotions, we look at them as not having a complete mind. Furthermore, if we consider someone else not to have a complete mind, and the ability to recognize and understand emotions, it is safe to assume that we also do not want this other one to make ethical choices on our behalf. If this is the case, then this consequence of not being able to make ethical choices obviously complicates the idea of algorithms taking up any leadership position. Leaders are expected to serve our interests and make the appropriate decisions to do so.
So, this should be the end of the story and algorithms should simply not move into leadership roles.
Right?
The leadership of today will not be the leadership of tomorrow
Maybe it is not the end of the story, but rather a new beginning. If we listen to popular media, business press and visionary keynotes, algorithms may still be in the leadership game – maybe even more than ever. Indeed, despite science identifying several important limitations that impede algorithms from taking up decision-making responsibilities, this has not prevented discussion on whether leadership by algorithm should still occur. The idea that algorithms can run organizations is not one that is dead and buried, but rather alive and kicking.
How then can this discussion about automated leadership survive and even be envisioned as the future leadership model? One possible reason may be that this leadership-by-algorithm hype in essence indicates a frustration with today’s (human) leadership. As a result, business and society at large may be looking for different forms of leadership. So, it seems likely that we are looking for a different kind of wisdom in our leaders of tomorrow. And, that kind of new wisdom could well be provided by an algorithm.
What I am saying is that we may have entered an era where we do not consider it necessary for our future business leaders to possess the kind of wisdom that we so dearly attribute to humans. Rather, it may be that we define the wisdom for our future leaders in terms of other attributes and skills. It could well be that we are looking for a kind of leadership that is best equipped to provide the most accurate and, at the same time, fastest decisions. If we want those decision-making qualities to be reflected in our future leaders, then it should not be a surprise that we are ready to embrace the idea of leadership by algorithm. After all, isn’t it the case that leaders able to make fast and accurate decisions should also be able to best manage a volatile business environment?
Interestingly, when we look at leadership literature, scholars in the past have portrayed good leadership as those who “make good decisions in a timely way.”57 We know that leaders have to make decisions on a daily basis. We also know that those decisions reveal important social consequences that can benefit or harm the organization and its employees.58,59 For these reasons, today’s focus in the digital era may be more on selecting leaders who are able to deal with data in the most optimal way. And, subsequently, we recognize suddenly the beauty of an algorithm as a likely candidate to make decisions and, hence, lead.
If we move from our theoretical exercise above and on to what we see in practice, we may find some evidence in favour of leadership by algorithm. The one thing that is not going unnoticed is that jobs are increasingly being automated, with algorithms integrated into decision-making processes. This trend could be interpreted as a signal that a new kind of automated leadership may well be on its way.
And, why should this be? Well, the faster acting, more accurate and consistent self-learning algorithms become, the more likely it could be that humans will gradually transfer