From Big to Democratic Data: Why the Rise of AI Needs Data Solidarity

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

2 Citations (Scopus)

Abstract

Datasets have come to play a significant role in the technical and political realities of our overdeveloped world. This chapter indicates how invisible data processes pose a threat to the health and safety of the global public and argues for the democratic potential of data practices. This potential is set to become even more influential due to the central role data plays for training contemporary AI and technologies such as machine learning. Our case study explores the role patient datasets have for machine learning research in healthcare and shows that publicly available datasets are central to advancing data analysis research; they can act as a counterbalance to datasets full of absences, biases, and disconnects that often corrupt the quality of data. Given this, we argue for the introduction of ‘data solidarity’ as a principle of data governance and an effective critical data practice that focuses on the democratic (instead of economic) potential of data; a potential that is far too often overlooked.
Original languageEnglish
Title of host publicationDemocratic Frontiers: Algorithms and Society
PublisherRoutledge
Publication statusPublished - 2022

Publication series

NameAlgorithms and Society
PublisherRoutledge

Fingerprint

Dive into the research topics of 'From Big to Democratic Data: Why the Rise of AI Needs Data Solidarity'. Together they form a unique fingerprint.

Cite this