Doctor of Philosophy
School of Humanities and Social Inquiry
The growing importance of big data in contemporary society raises significant and urgent ethical questions. In the academic literature and in the media, the dominant response to many of these ethical questions is to re-examine the role and importance of privacy protections, but I argue that it is far more fruitful to investigate the relationship between power and big data. As algorithmic processes are increasingly used in decision-making processes, it is crucial that we understand the ways in which big data can be used as a technology of power. Only then can we properly understand the ways in which the use of big data impacts on and reorganises society, and go on to develop effective, tailored protections for individuals against harm from the use of big data. First, I show that the rise of big data highlights the limits of privacy protections, as big data-based analytics allow for personal information to be inferred in ways that circumvent privacy protections and problematises the category of personal information. In order to properly protect people from the potential harms that can arise from the use of big data in decision making, I argue that we must also examine the relationship between big data and power. In this thesis, I will present an argument for a pluralistic understanding of power, and a lens through which we can identify the kinds of power being exercised in the contexts we are investigating. Power is best understood as an umbrella term that refers to a diverse range of phenomena across an equally diverse range of domains or contexts. We can use this attitude to examine the central features of an exercise of power to identify the relevant theoretical accounts of power to draw on in understanding the modes of power present in a context. In Chapter 4, I will demonstrate the value of this approach by using it to analyse four contexts where big data is used as a technology of power, showing that we cannot use a single theoretical understanding of power across all exercises of power. Following this, I examine the impacts of big data on the operation of power. While many in the literature see big data as necessitating the development of new theoretical understandings of power, I argue that there are important historical continuities in power. Big data can be picked up and used as part of existing kinds of power just as any new technology can, and while this may change the efficiency, range, and effectiveness of exercises of power, it does not change their fundamental nature. However, there are impacts on the operation of power that are unique to big data, and one of these impacts I consider here is that the inferential capabilities of big data shift power from acting on human subjects and towards acting on data doubles (fragmentary digital representations of people). This leads to significant ethical problems with ensuring that power is exercised accountably. Finally, I will demonstrate these problems in Chapter 7 through examining four more contexts in which big data is used as a technology of power, showing how the shift to the data double as the subject of power undermines the effectiveness of accountability as a check on the abuse of power.
Reynolds, John, Big Data as a Technology of Power, Doctor of Philosophy thesis, School of Humanities and Social Inquiry, University of Wollongong, 2021. https://ro.uow.edu.au/theses1/1361
FoR codes (2020)
5001 Applied ethics, 500103 Ethical use of new technology, 500201 History and philosophy of engineering and technology
Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.