Читать книгу Leadership by Algorithm - David De Cremer - Страница 8

Chapter 2: The Leadership Challenge in the Algorithm Age

The machine age arrived a long time ago, but today’s need for the machine seems to know no limits. Modern machines need more room, more execution power and yes, maybe also the desire to lead. But who will they lead? The answer is, those who need the machine the most. And, dear reader, this may well turn out to be humans.

Naqvi writes that “the need of the machine could not be fulfilled without getting the followers to co-operate.”36 As we will see later, the influence that leaders achieve – and which we will call the process of leadership – can only be accomplished if there are others out there willing to follow the directives, ideas and suggestions of the one leading. So, if machines are about to become leaders, they need followers to achieve the powers and potential that we attribute to those same machines. As humans have developed an obsession to ensure optimal use of the potential attributed to machines, it follows that the followers are likely to be those same humans. It is a string of thought that keep the minds of scholars today very occupied.

Of course, many among us may wonder whether a machine (empowered by the workings of algorithms) leading the human species is actually a valid proposition. Why should we even reflect on the possibility of algorithms leading organizations populated by humans? Does it make sense to ponder whether your next boss will be an algorithm? What should we do about it? Does ‘Leadership by Algorithm’ even have a basis to exist? And, if so, do we really need it?

To delve deeper into this series of questions, we first need to ask ourselves whether it is a reasonable thing to expect human employees to follow algorithm-driven leaders in the same way they would a human leader? Can such a world exist? Some scholars think so.

These scholars assume that whoever becomes the leader is determined, to a large extent, by the situation at hand. One of the most prominent leadership scholars, Jeffrey Pfeffer, supported this view in his 1977 Academy of Management Review piece, ‘The Ambiguity of Leadership’. In his article, Pfeffer debunks the myth that leaders are a unique kind of species – independent of any situational influence. Pfeffer argues that in our human drive to see heroes as the true leaders, we adopt the illusion that only those individuals who bring something special to the table can be called leaders. (As a side note, when looking at the contemporary movie industry, with its focus on the Avengers franchise and other action heroes, it is clear that today we still have a need for the illusion of grand and heroic leadership.)

But, interestingly enough, it is actually the other way around; the situation makes the leader. And history supports this, revealing examples where leadership is attributed to those who, for reasons other than their own unique capabilities, win wars (Sun Tzu’s The Art of War), or can give the impression that their office is doing wonders for the economy (President Donald Trump bragging that ten years without recession is his own doing).

One of the most extreme examples of this in my lifetime is President George W. Bush and the tragedy of 9/11. Before the Twin Towers and the Pentagon were hit by hijacked airplanes, Bush had some of the lowest leadership ratings since records began. In the aftermath of these events, he visited Ground Zero and announced that the US would punish those responsible. And something extraordinary happened. Suddenly, a man considered by most as incapable of taking on the role of President of the United States was elevated to one of the highest ratings for leadership ever recorded. The situation caused Bush to be seen by many as a good leader, especially when he expressed aggression and optimism, and took the actions of a leader going to war.

The power of the situation to decide who will lead has been proven by various scientific studies of mayors, athletic coaches and corporate leaders.37,38 These insights have led to the conclusion that if the situational demands dictate leadership effectiveness in the eyes of others, then “what does it matter who occupies the position or how they behave?”39

Will your boss be a robot?

These acclaimed insights seem to suggest that it is OK for algorithms to take up the role of leader. Just as it is OK for humans to take on this role. So, if this is true, why worry about automated leadership?

If it does not matter, then, as humans, we should take even more responsibility to ask whether algorithms are really able to lead organizations. Today’s organizations are faced with a volatile business environment and are therefore required to act in fast and agile ways. To meet these demands, humans seek to explore how technology can help us to operate more efficiently and manage optimal performance. For instance, technological innovation is needed to ensure that our organizations can adapt to deliver products and services to a market populated with demanding customers. There is no way to escape this truth. It is the situation that we are faced with!

Our business environment demands algorithms to be part of promoting organizational efficiency. Knowing the power of situational influences, it may not seem such a crazy idea at all that those running our organizations in the pursuit to excel may well become dictated by algorithms. Or, in other words, algorithms may well drive the management process very soon.

This is not simply a thought exercise anymore. Being confronted with greater expectations of productivity, the need to respond faster, and the requirement to be more rational and data driven in our responses, both business and thought leaders have put the idea to automate leadership firmly on the table. Few are now questioning if this is feasible, instead they are wondering how best to implement their AI management strategy. Business leaders have embraced the idea that the widely announced digital disruption has introduced many challenges. Maybe too many for humans to deal with. As a result, it has made the business world uncertain on how to manage digital disruption. If this is the case, should it then not be better to rewrite the leadership handbooks that we have written over the last few decades?

Our organizations have become so complex that running them seems to require a leader with almost superhuman abilities. The world moves too fast for human leaders to be confident that they will make good decisions. Even though human intelligence may be a beautiful and complex thing, the simple reality is that it is not always up to the job of processing massive amounts of data, in very short periods of time, to arrive at the best decision available. And, this is where the algorithm jumps into the leadership equation. Algorithms can replace almost any feature that we think of as representing good leadership. We have arrived at a time where, “for any given skill one can think of, some computer scientists may already be trying to develop an algorithm to do it”.40 As a result, it is not a utopian idea any more to point out that the future management of organizations will be an automated one.41 It will prove to be cheaper, more efficient, and potentially more impartial in its actions than human beings.

I would even go further: Given the market demand for business efficiency, it is inevitable that algorithms are and will be replacing various jobs at different levels of managerial discretion and as such seem destined to take over leadership from humans in many ways.42 As Frank Pasquale noted, “authority is increasingly expressed algorithmically”.43 In fact, having algorithms as part of how authority looks and sounds, moves the algorithm by definition into the role of power holder. How so? Think about it, because of their deep learning abilities, algorithms construct decision rules that are not well understood by humans. Their decisions will therefore feel fixed and in some way intractable. If we thus allow algorithms to take up the role of the authority, we will become dependent on them. In other words, if we voluntarily make algorithms part of our leadership system, then we create a context where we will be dependent on them. In addition, this dependency will – because of the rational nature of an algorithm – feel distant, rigid and hard to argue with. The result will be that algorithms will co-ordinate what we do and how we do it.

The red-hot business model

If algorithms have the power to shape our interactions, how does the corporate world factor this into their ideas on how to run a company? Well, it seems increasingly likely that they are embracing a new business model, where algorithms lead in the management of decisions and execution of jobs, and human employees follow.44 For example, the analytics provider SAS considers digital data management as a step forward, where algorithms not only provide advice but also fuel strategic decision-making.45 This business model portrays human employee’s bosses as the smart ones, since they are driven by algorithms that can perform in superior ways. No doubt this business model is clearly present in the minds of many as it drives important leadership transformations currently taking place in organizations.

Efforts are underway to understand the brains of successful CEOs and use those neural imprints to create highly efficient algorithms.46 These algorithms will continuously learn and develop into a leading authority, superior even to human leaders. Results from a survey conducted by the World Economic Forum’s Global Agenda Council on the Future of Software and Society also revealed that businesspeople expect artificially intelligent machines to be part of a company’s board of directors by 2026. As a matter of fact, this futuristic view is already materializing. For example, the Hong-Kong based venture-capital firm Deep Knowledge recently appointed a decision-making algorithm – known as VITAL – to its board of directors, indicating that algorithms are already taking on the leadership challenge to set parameters of corporate governance.47,48 And, last, but not least, in 2019, Amazon allowed AI to fire employees without consulting with any human.49

It is important to realize, however, that there is also the potential for all of these optimistic and exciting developments to backfire. They could create a situation where people are increasingly being confronted with existential doubts and fears about their vision of the future. Indeed, although algorithms bring economic benefits, people may feel that human labor is devalued by organizations aggressively pursuing automation.50 As a result, the value of being a human employee is not understood very well anymore and the fear for unemployment is a real one.

In addition, existential doubts may create the need to reflect on what kind of society we want to see. Do we want a society where the corporate dream of having optimally functioning organizations leads us towards automation of those leading us? Or do we want a society where we decide not to forego the human touch in whatever we do, including leadership?

Such feelings cannot be underestimated because they directly link to many people’s uncertainties about whether their job will still be relevant in the new technology era. Many also question what the future of human employees will be if algorithms run the decision-making process. It is these uncertainties that I, as a business school professor, am faced with when executives ask whether leadership courses, which provide insight into human motivation, will disappear in the future. This is the point where I see it is necessary to disrupt our thinking about the business models we want to adapt and the kind of automated leaders we want to see in the future.

It is all well and good to have these models in mind. Its exciting, even, to see where the limits of those models may lie in our pursuit to optimize performance, organizations and society. However, it is also the responsibility of humans to be critical about their own ambitions, desires and wildest dreams. Because, what one can imagine is not necessarily what we need, nor is it necessarily the vision that is driven by the best and most accurate information. In fact, alongside all the exciting technological developments that we witness today, when it comes down to automating our business leadership, we need to realize that those wishes could well be driven by people who are poorly informed about the real impact of automated work forces.

Human sophistication

So, the evolving business model of the future seems to be one designed and pushed by people who do not necessarily have the required knowledge of what algorithms are capable of, nor of what kind of human skills are necessary to drive leadership excellence.

It is a fact that many business leaders cannot be recognized as experts in technology, its applications and usage, and philosophical thought regarding the human reality that develops in an automated environment. So, despite all the greatness and beauty behind the idea that increasing automation will inevitably lead to automation of everything, including the leadership of organizations, we also have to be critical about what exactly has real value and what has not. In this respect, it is interesting that, contrary to the preferred automation model of business leaders, recent research has revealed that skills related to feelings will define the future jobs for humans. In fact, salaries for human employees in the future are expected to be determined more by the ability to deal with emotions and relationships rather than by their cognitive abilities. This reality paints a future where jobs that require sensitivity to needs for relationships will have to be populated by humans and the role of leadership seems to fit that bill.

The argument that I am putting forward is that the functioning of our organizations and societies are not served by a kind of sentiment that the analysis of data by algorithms will automatically develop and lead strategies in miraculous ways. Algorithms are not technological tools that have the leading abilities to deliver immediate returns without any human presence or interference needed. As we see technology develop today, we need to be aware of the fact that automated decision-making is still something of a black-box that runs in less structured ways than we think. Algorithms also miss human sophistication, and an awareness of moral norms and emotions; all skills that allow leaders to create value beyond the immediate observable financial returns. In fact, when looking at the data available, reality paints a somewhat different vision when it comes down to the optimal use of algorithms in leading and co-ordinating organizations.

Research by IBM shows that 41% of CEOs report that their organization is not at all prepared to introduce data analytic tools into their management structures.51 In addition, when it comes down to dealing with humans in automated ways, only about 22% of organizations say that they have adopted algorithms in their Human Resources practices.52 And, of those 22%, most are not clear on what the exact effect is that the algorithms reveal. Given these numbers, it seems reasonable to argue that wise leadership in the 21st century will still need more of a strategy than simply trying to make a difference by means of optimizing the technology (including the management technology) taking care of our data. Rather the real difference will be made in having leadership out there that can make use of these technologies in human-centred and sustainable ways that benefit human values, interests and well-being.

36 Naqvi, A. (2017). ‘Responding to the will of the machine: Leadership in the age of artificial intelligence.’ Journal of Economics Bibliography, 4(3), 244-250.

37 Gamson, W.A., & Scotch, N.A. (1964). ‘Scapegoating in baseball.’ American Journal of Sociology, 70, 69-72.

38 Pfeffer, J., & Salancik, G.R. (1978). ‘The external control of organizations: A resource dependence perspective.’ New York: Harper & Row Publishers.

39 Pfeffer, J. (1977). ‘The ambiguity of leadership.’ Academy of Management Review, 2, 104-112.

40 MacCrory, F., Westerman, G., Alhammadi, Y., & Brynjolfsson, E. (2014). ‘Racing with and against the machine: Changes in occupational skill composition in an era of rapid technological advance.’ In Proceedings of the 35th International Conference on Information Systems (pp. 295–311). Red Hook, NY: Curran Associates Inc.

41 von Krogh, G. (2018). ‘Artificial intelligence in organizations: New opportunities for phenomenon-based theorizing.’ Academy of Management Discoveries, 4(4), 404-409.

42 Parry, K., Cohen, M., & Bhattacharya, S. (2016). ‘Rise of the machines: A critical consideration of automated leadership decision making in organizations.’ Group & Organization Management, 41(5), 571-594.

43 Lindebaum, D., Vesa, M., & den Hond, F. (in press). ‘Insights from the machine stops to better understand rational assumptions in algorithmic decision-making and its implications for organizations.’ Academy of Management Review.

44 Derrick, D.C., & Elson, J.S. (2019). ‘Exploring automated leadership and agent interaction modalities.’ Proceedings of the 52nd Hawaii International Conference on System Sciences, 207-216.

45 SAS (2018). ‘Becoming a data-driven organization.’ https://analyticsconsultores.com.mx/wp-content/uploads/2019/03/Becoming-a-data-driven-organization-Citizen-Data-Scientist-SAS-2018.pdf

46 Copeland, R., & Hope, B. (2016). ‘The world’s largest hedge fund is building an algorithmic model from its employees’ brains.’ Retrieved from https://www.wsj.com/articles/the-worlds-largest-hedge-fund-is-building-an-algorithmic-model-of-its-founders-brain-1482423694 on 31 October 2018.

47 Nelson, J. (2019). ‘AI in the boardroom – Fantasy or reality?’ March 26. Retrieved from http://www.mondaq.com/x/792746/new+technology/AI+In+The+Boardroom+Fantasy+Or+Reality

48 Libert, B., Beck, M., & Bonchek, M. (2017). ‘AI in the boardroom: The next realm of corporate governance.’ February 21. Retrieved from https://sloanreview.mit.edu/article/ai-in-the-boardroom-the-next-realm-of-corporate-governance/

49 Amazon (2019). https://www.businessinsider.sg/amazon-system- automatically-fires-warehouse-workers-time-off-task-2019-4/?r= US&IR=T

50 Acemoglu, D., & Restrepo, P. (2019). ‘Robots and jobs: Evidence from US labor markets.’ Journal of Political Economy. Accepted August 1.

51 IBM (2019). ‘Unplug from the Past: 19th Global C-Suite Study,’ IBM Institute for Business Value, 2018, https://www.ibm.com/downloads/cas/D2KEJQRO

52 LinkedIn (2019). ‘The Rise of HR Analytics,’ 2018, https://business.linkedin.com/content/dam/me/business/en-us/talent-solutions/talent-intelligence/workforce/pdfs/Final_v2_NAMER_Riseof-Analytics-Report.pdf.