lead to those same algorithms. Today’s reality is that companies operate in complex and volatile business environments where a need for faster and more accurate decision-making is increasingly emerging. To be able to deal with this need, people begin to seek a new kind of leader to get the job done. This new kind of leader seems to be recognized as one that is automated, thinks rationally and, hence, offers a sense of accuracy and precision to provide the most optimal choice in any kind of situation.
Are we ready to accept?
At this moment, you may wonder whether our need and desire for this kind of new leader is not simply wishful thinking. On one hand, humans surely would not so easily transfer the power and influence of their leaders to an automated entity. On the other hand, maybe we are too sentimental about these questions. It is this emotional reaction that prevents us from turning leadership as we know it today into a more optimal form that may well be less human.
After all, from a rational point of view, we should do anything possible to optimize our way of doing business. Such efforts, without a doubt, should also include thinking about how we can run our organizations and make decisions in better, more optimized ways. If we are serious about thinking in a rational fashion about how we want to approach the future of work, it should only be a matter of time before automated leadership will happen.
Some research actually suggests that people today are ready to accept this idea. A 2019 study by Logg, Minson and Moore examined the attitude of humans towards the judgments and advice offered by algorithms. They arrived at some powerful conclusions.60 The authors concluded the following: “Our studies suggest that people are often comfortable accepting guidance from algorithms, and sometimes even trust them more than other people. … It may not be necessary to invest in emphasizing the human element of [the] process. Maybe companies that present themselves as primarily driven by algorithms, like Netflix and Pandora, have the right idea.”
These recent studies clearly underscore the idea that humans may not be so sentimental about the future leadership question and are actually more willing to trust the input and direction that algorithms provide than we expected. And, importantly, this tendency does not seem to be the case only in the context of experimental studies, but also in real life. We know that Uber riders respond in a more negative way to a price increase if it is set by a human, compared to when it is set by an algorithm. If a human sets out new guidelines that harm the self-interest of other humans, then those others will be less forgiving than when an algorithm initiates this policy.
After all, if a decision made by a human leads to negative consequences for another human, then it is usually evaluated as intentional. Whereas, when an algorithm makes a decision that leads to negative consequences for a human, people do not perceive the algorithm to have acted intentionally. Indeed, as discussed earlier, people do not perceive an algorithm to have a mind and therefore it is difficult to see how an algorithm could have bad intentions.
This shows that, under certain circumstances, humans actually prefer algorithms to make choices because they are not considered to be threatening to us. A recent study published in Nature Human Behaviour consolidated this idea.61 Studies revealed that human employees showed a stronger preference for being replaced by an algorithm than by another human. This is quite a surprising finding in light of the current debate that people are fearful of being replaced by AI. But then again, from the point of view that people may be more lenient toward having rationally-acting algorithms to make decisions on their behalf, it may not be that surprising. Researchers in this study found that the reason behind their finding was that humans experienced being replaced by another human to be more harmful to their self-interest. To be more precise, they considered being replaced by another human to be more threatening to their public image (how they are perceived by others) and self-esteem.
In fact, if you are replaced by another human, people may quickly reason that this other person is better than you. This judgment is something people obviously do not like, because it implies that others will have a negative view of you and your abilities (i.e. which is part of your public image). On the other hand, if you – as a human – are being replaced by an algorithm, you are being replaced by a non-human, and this event is experienced as less threatening to your public image. Indeed, a human and non-human are completely different species and therefore cannot be compared in the same dimension.
Both the Uber and research findings in Nature Human Behaviour illustrate circumstances where people prefer an algorithm to be in charge, as opposed to a human. Thus, these observations emphasize that humans are more inclined to rely on the actions and advice offered by algorithms exactly because they are non-human and, as such, take out the emotional and biased side of human decision makers. It may well be the case that humans are getting ready to see the benefits of a rationally-acting non-human machine to replace those positions that need more optimal and responsible decision making. Leadership positions, for example.
If our leaders do not let their intentions and biased judgments play a role in the decisions they take, we, as humans, will take it less personally and will be more assured of the decision being fair. From that point of view, we see that algorithms are gaining ground quickly in the area where they can provide advice to customers and act as experts to higher-level managers in organizations. For example, recent surveys among customer support and service operations report up to 85% willingness to use virtual assistants by 2020, leading to expectations that in the near future we will witness a significant increase in chatbot use by companies.62,63
Who qualifies?
What to think of all of this? Well, if we take into account the fact that businesses today operate in volatile and complex business environments, and therefore require fast and optimal decisions; the recent scientific evidence; and the optimistic vision of business people that automation will hit all levels of authority, then we can only conclude that leadership by algorithm is the best way forward.
It will happen and apparently for good reason. In fact, online work environments are becoming increasingly popular, and the application of algorithms to monitor, co-ordinate and evaluate performance of employees in these settings is very much present.64 Even more so, businesses today admit that they are increasingly relying on algorithms to co-ordinate work relationships. Delegating the power of authority to algorithms definitely does not seem like science fiction anymore. But is this really the end of the story? Is the final conclusion that leadership by algorithm will happen and that it will be for the better?
In my view, it is not! Yes, it would be the end of the story if we define leadership in a narrow way. So, what is meant by a narrow manner? To answer this question, let us first look at how we are approaching the whole issue of replacing humans with algorithms. Many people, and especially the popular press, only look at the skills that are required to execute the job. As we all know, however, doing a job well entails more than simply being able to tick the boxes representing a list of skills. Of course, skills are important and recognized as relatively good predictors of how well employees will perform. But that is not the only thing we need to become influential and effective at work. Another important aspect to ensure effective job execution is that meaning is given to the job. For people to stay motivated in their job, it is crucial that their function is understood in the broader organizational setting.
Jobs are part of a broader social context in any organization. And it is because of this broader social context that employees are also required to possess the social skills to talk, negotiate, lobby and collaborate with others. Unfortunately, it is also this element of giving meaning to the job in a broader work environment that is hardly ever a focus in the discussion of whether or not jobs should be automated. I argue that we are facing the same problem when we are talking about whether algorithms should and can move into leadership roles.
In today’s discussions, a trend has emerged that leadership is only looked upon as a set of required skills. If all the boxes are ticked, a person should be ready to assume a leadership role. The consequence of looking at the possible automation of leadership in this way is that organizations are too narrow in their thinking about what it takes for automated systems to run the organization. Specifically, this rather narrow way of defining leadership means that organizations will make the simple calculation that if any actor (human or non-human) delivers the skills needed to make decisions in fast and data-driven