Why can’t you buy yourself a Steve Jobs? Part 1

Posted by

Have you ever wondered why you can’t buy a manager for your company? It’s not about hiring a human manager, but buying a machine manager of a certain type, such as one who manages the production or marketing department. Making such a manager digitally shouldn’t be so difficult after all, and yet… there is no manager store!


On the one hand, there are a number of phenomena that should cause one to eventually be able to buy oneself a Steve Jobs or Elon Musk. First, Increasing digitization and automation is a dominant feature of the modern world. During the last 20 years there have been a time of extremely rapid advances in digital technologies, artificial intelligence, machine and deep learning, etc. Managers are increasingly using electronic tools that make it possible to record their work, and indirectly record the performance of the organization [1].

Secondly, for several years now, there has been a growing number of publications on the vision of replacing the manager with computer software, and as a result, creating robot managers. Admittedly, these are not typical humanoid robots with which one can talk about a salary increase or resolve a conflict with another employee, but tools to support decision-making, problem-solving or analysis of a phenomenon [2].

On the other hand, when it comes to replacing the human manager with a robotic manager, as early as 1967 Peter Drucker wrote that computer systems (at that time – computers) will not only be used to collect information, but the algorithms stored in them will in time be able to replace the work of managers [3]. Despite of several decades, this has not happened, although computer systems are filling and automating more and more areas of human life and, consequently, the work of the manager. The question can then be asked, why can’t a robot be hired as a manager in the third decade of the 21st century?

It seems that there are several conditions that have not yet been sufficiently met for this to be possible. First, the reason why it is difficult to build an algorithm that performs the tasks of a manager is the unpredictability of the behavior of the entities that interact, in this case the robot manager and the participants of the organization [4]. It is possible to create stochastic or deterministic models in this regard, but their accuracy is low and it is difficult to talk about activating adequate behavior of the robot manager.

Secondly, there is the problem of how to generate the possibility of real influence of the robot manager on team members and vice versa [5]. It turns out that exerting influence does not only refer to giving verbal orders. It also refers to the organizational culture created by the manager, his non-verbal behavior, the closeness in the relationship with the employee or charisma, as this boss possesses.

Thirdly, it is important that there is a common basis for the robot manager and team members to communicate knowledge about organizational reality (however, this is not just about the language used to communicate, but also a common conceptual framework and meaning of concepts) [6]. This is perhaps the most difficult to deal with when it comes to natural language and understanding it through man-made algorithms. High contextualization is not conducive to the human-machine relationship and severely hinders understanding each other.

What other obstacles are there that we can’t build a copy of Steve Jobs? They lie in management science itself and the way we study what a manager actually does. You’ll read about them in the next post on this blog.

[1] B. Ewenstein, B. Hancock, A. Komm: Ahead of the Curve: The Future of Perform- ance Management. “McKinsey Quarterly” 2016, May. http://www.mckinsey.com/business-functions/organization/our-insights/ahead-of-the-curve-the-future-of-performance- management.

[2] M. Chui,  J. Manyika, M. Miremadi: How Many of Your Daily Tasks Could Be Automated? “Harvard Business  Review” 2015, December. https://hbr.org/2015/12/how-many-of-your-daily-tasks-could-be-automated; Eidem:  Four Fundamentals of Workplace Automation. “McKinsey Quarterly” 2015, November. http://www.mckinsey.com/business-functions/business-technology/our-insights/four-fundamentals-of-workplace-automation; T.H. Davenport, J. Kirby: Beyond Automation. “Harvard Business Review” 2015, June. https:// hbr.org/2015/06/beyond-automation; M. Dewhurst, P. Willmott: Manager and Machine: The New Leadership Equation. “McKinsey Quarterly” 2014, September. http://www.mckinsey. com/globalthemes/leadership/manager-and-machine; D. Fidler: Here’s How Managers Can Be Replaced by Software. “Harvard Business Review” 2015, April. https://hbr.org/2015/04/heres-how-managers-can-be-replaced-by-software#

[3] P.F. Drucker: The Manager and the Moron. “McKinsey Quarterly” 1967, December. http://www.mckinsey.com/business-functions/organization/our-insights/the-manager-and-the-moron

[4] G. Klein et al.: Common Ground and Coordination in Joint Activity. W: Organizational Simulation. Red. W.R. Rouse, K.B. Boff. New York, John Wiley & Sons, 2005.

[5] K. Christoffersen, D.D. Woods: How to Make Automated Systems Team Players. “Advances in Human Performance and Cognitive Engineering Research” 2002, 2, s. 1–12.

[6] H.H. Clark, S.E. Brennan: Grounding in Communication. W: Perspectives on So- cially Shared Cognition. Red. L.B. Resnick, J.M. Levine, S.D. Teasley. Washington, American Psychological Association,  1991.