The development of artificial intelligence (AI) enables broader automisation and the redefinition of a large number of professions. One not-so-talked-about area is the tasks that are traditionally performed by managers. This very subject was, however, the topic for discussion at a seminar that was held by Unionen and AI Innovation of Sweden.
The managerial role as we know it today involves a broad palette of tasks, many of which could be automated using AI. One of these tasks is administration. Another is the collation, storing and analysis of data that can be used to further optimise the task of management. The managerial role also involves responsibility for making decisions. This includes responsibility for the data and competences that are the basis for those decisions and for taking action. In this context, we should be talking about a so-called "algorithm-based management".
Victor Bernhardtz, digital labour markets at Unionen, stresses the importance in his work of trying to understand the impact of digitalisation on labour, labour planning and how it affects the work/society relationship. Earlier this year, Unionen published Victor's report "Machines as bosses - About cognitive data systems, algorithm-based labour management and the Swedish labour market.”
At a morning seminar hosted by Unionen and AI Innovation of Sweden on 30 September, Victor presented further material from his study, showing how the labour market will be affected by new technologies, with a particular focus on the role of managers.
"Relatively soon, we will probably see an increase in algorithm-based management in various areas of the labour market; white collar and blue collar", tells Victor.
"One example might be the use of data systems to fully or partially automate HR functions like recruitment. We may also see decision-making tools for bosses, such as full or partial automisation of required labour assessments."
"A desire for management that is driven more by data means increased data requirements. These bigger opportunities bring bigger challenges, not least with regard to ethics."
"The single most significant change that I foresee is that more and more companies and organisations will want to strengthen their capacity for working in a way that is driven by data. By this, I mean that they will increasingly analyse large amounts of data in order to produces a more precise basis for making decisions. This might be data about production volume or customer consumption trends, or lots of other things. A desire for management that is driven more by data means increased data requirements. These bigger opportunities bring bigger challenges, not least with regard to ethics."
"I still think that most managerial tasks will end up being performed by people. Ultimately, I do not think that machines will be able to replace people in management. Machines may be able to perform deep analysis, but it will be people who will need to interpret that analysis and show the way forward."
Algorithm-based management has a lot of potential in terms of quality and efficiency. Victor points out, however, that there are various aspects that we perhaps need to be aware of."
"One concern with algorithm-based management is that people will rely too heavily on the conclusions and recommendations of cognitive computer systems without a critical analysis, such as an assessment of bias in the data that is used, or not having the full picture."
"It is, for example, hard to imagine that a computer system will have access to all the data that it would need to do a thorough analysis of a person's potential, abilities or work performance. We need to be aware of the limitations."
The role of the manager is not just about decisions and data, but entails a relationship with the employee. This area is one of those that Victor thinks will be most difficult to replace.
"Complex human interactions are, I would say, something that machines are not going to be able to learn in the forseeable future, and in particular the relationship between manager and employee.
At present, we see examples where unsuccessful attempts have been made to automate managerial tasks. Often this is because the automation has eliminated the ability to interact with a manager, which forces the individual to integrate with "the system". There is a risk that people will not receive the right leadership, support and feedback. I'm not saying that such solutions couldn't work, but personally I have yet to see one that I am satisfied with."