Back to search

University of York

Not Content: How the algorithmic telos cultivates radical political outcomes by its recommendation of media

Abstract

dc:description.abstract

This paper investigates a gap in prior research into algorithmic recommendation. Namely, the precise interactive mechanism between the corporate end-goal of user retention, and the outcomes that eventuate from it, including radicalisation and violence. I show that both premises have been established by prior research and explore how the formal traits of media on algorithmically curated platforms that maximise user retention also lead to ideological extremism. I achieve this through analysing patterns in recommendation using an experimentally generated dataset of algorithmically autoplayed media as proxy. I track the suggested videos of YouTube accounts with a range of simple existing media habits to account for the impact of pre-existing political preferences that have been the focus of much of the existing literature. Therein, I find that more universal factors drive recommendation. Results indicate that while there are clear correlations in the formal factors, the actual content of recommended media develops erratically and with little evidence of a linear progression towards politically radical outcomes. Instead, recommendations follow patterns of type, with a continuity of genre, and user demographics especially, with little coherence in the actual topic. Promoted media share a number of apparently algorithmically privileged formal factors — notably runtime, sensationalism, misinformation and niche — which I reason are also formal factors shared disproportionately by radical reactionary content. My research thus demonstrates the formal factors discussed in my thesis that encourage increased user retention above all else are also those associated with extreme content. Industry attempts to address platforms’ radicalisation pathways from algorithmically-driven content with post hoc content moderation is inadequate. In combination with prior literature, my findings suggest that social media recommendation motivated by this telos of retention maximisation for profit pushes users toward media with the formal factors and impact on the user that radical content has.

Degree

thesis:*
Name dc:type.qualificationname
MA by research
Level dc:type.qualificationlevel
masters
Grantor dc:publisher.institution
University of York
Year dc:date.issued
2024

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Tomkins, Oliver
Advisor dc:contributor.advisor
  • Ng, Jenna

Rights

Language dc:language.dcterms_iso639_2
eng

Chain of custody

source
Harvested from
White Rose University Consortium
Base URL
etheses.whiterose.ac.uk/cgi/oai2
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
related terms
citation

Tomkins, Oliver. Not Content: How the algorithmic telos cultivates radical political outcomes by its recommendation of media. masters thesis, University of York, 2024.