Freedom Under Algorithms: how unpredictable and asocial management erodes free choice

Robert Donoghue^1,2*

• ¹University of Essex, Colchester, United Kingdom
• ²School of Business, Faculty of Social Sciences, University of Essex, Colchester, Essex, United Kingdom

This article examines the impact of algorithmic management on individual freedom. To orient this exploration, I draw on the (feminist) conception of liberty as the choosing subject. The central suggestion is that algorithmic management poses a serious threat to an indispensable part of the freely choosing subject: namely, it degrades the ability of subordinates to reasonably foresee the consequences of their choices and consequently, fully realise their personality. I call this phenomenon the 'foresight endangerment problem' and argue that it has both a technical and a social face. The technical face highlights the inherent unpredictability of advanced algorithms, including those that execute managerial functions. This issue is further complicated by the fact that as algorithms become more resilient and useful, their outputs grow increasingly opaque and unpredictable -what some refer to as the resilience-predictability paradox. The technical face is made manifest in the reported experiences of workers in the gig economy who describe experiencing unpredictable managerial decisions that they cannot anticipate nor easily contest. Subjection to such managerial randomness erodes their ability to make informed choices in service of their personal goals. The social face emphasises the consequences of disembedding managerial power from social relationships between humans to asocial relationships between humans and software. Subordinates of human managers enjoy a vast number of tools to predict managerial thinking that arise from the intricate and complex processes of social interaction. The disembedding process forecloses the use of these tools and fundamentally undermines the capacity of subordinates to promote their ends through free choice.