skip to content

Sociology Research


A project led by Dr Ella McPherson examines whether the significant trend in the digital age is the rise of autonomous systems, which use big data to inform algorithms through machine learning.  One significant consequence of this trend has been the expansion of these autonomous systems into areas previously the preserve of human judgment and decision-making.  For example, autonomous systems are at work parsing documents in the discovery process of law; algorithms helped journalists comb through the Panama papers; and Google is working with the NHS to support doctors with predictions based on patient big data. 

Once considered the preserve of human experts, valued for their judgment and decision-making, practices in law, journalism, and bio-medicine are increasingly being delegated to machines.  These machines’ autonomous systems are doubly opaque; first, citizens often have no insight into the algorithms governing them; and, second, because these algorithms are not explicitly programmed but instead continuously evolve due to machine learning, citizens and even the algorithms’ programmers have little control over their changing formulas.  This shift clearly raises ethical and governance challenges to existing practices – challenges which urgently must be understood as this shift becomes more prevalent and pervasive across sectors.

This project builds on the existing interdisciplinary work of the Ethics of Big Data CRASSH Research Group to conduct seed research and networking for the formulation of a large grant application on the topic of ‘Ethics and governance of autonomous systems in the digital society.’  This large grant would significantly expand the scope of current research into ‘algorithmic accountability’ by investigating the implications for ethics and governance of how machine learning methods and algorithms are integrated into wider computational systems and into social processes of decision-making and accountability more generally.

Image Credit: 'Big_Data_Higgs by KamlPhuc [CC BY-NC-ND 2.0].