Benjamin Jacobsen
Lecturer

Profile

Biography

I joined the department as Lecturer in 2023, and my research broadly examines the ethicopolitical effects of data, social media platforms, and machine learning algorithms on society and culture. I did my undergraduate degree in English Literature (2013-2016) and my MA in Society, Culture and Globalization (2016-2017), both at the University of York. In 2020 I completed my PhD, supervised by Prof David Beer and Dr Daryl Martin, with the project titled ''You Have a New Memory': Mediated Memories in the Age of Algorithms.' In 2021 I became a postdoctoral research associate on the ERC-funded project 'Algorithmic Societies: Ethical Life in the Machine Learning Age,' led by Prof Louise Amoore at Durham University.

In addition, I am the principal investigator on the ERC-funded project 'The Ethics of Synthetic Data in the Age of Machine Learning and AI' (SYNDATA, 2026-2030), and an LCAL Associate at the Leverhulme Centre for Algorithmic Life at Durham University. 

Departmental Roles

• Departmental Employability Lead

Research

Overview

My research broadly examines the ethicopolitical effects of data, social media platforms, and machine learning algorithms on society and culture. I am currently working on three principal research strands: firstly, I examine the emergence of synthetic data and how they are reconfiguring the condition of possibility for machine learning and ideas of ethics; secondly, the intersection of algorithms, social media, and memory in everyday life; finally, how algorithmic systems are transforming visual culture as well as what is made visible.

Projects

SYNDATA (2026-2030)

I am currently Principal Investigator on the five-year ERC Starting Grant project titled 'The Ethics of Synthetic Data in the Age of Machine Learning and AI' (2026-2030). The project aims to provide cutting-edge social science knowledge on how the emergence of synthetic data is transforming the relationships between data, AI, and ethics in our algorithmic societies.

The project aims to address three main research objectives, each of which is designed to address a crucial aspect of the ways in which the intersection of synthetic data, machine learning, and AI are transforming society. These aspects are: 1) extraction and generativity; 2) diversity and difference; 3) bodies and resistance.

Objective one: Extraction and generativity – To understand the fundamental relationships between algorithms, real data, and synthetic data.

The focus of objective one is to better understand the relationship between synthetic data and ‘real’ data. AI models such as GPT-4 and Midjourney are transforming how society sees data outputs such as images or text. While noted for their capacity to generate novel and realistic outputs, there has also been a social and political backlash. This backlash is rooted in the dual claim that these systems fundamentally rely on the extraction of people’s data without their consent and, on the other hand, that these are not novel outputs but instead only derivative of the real-world data the model was trained. This therefore raises the question to what extent synthetic data are simply ‘derivative’  or ‘proxies’ of the real data on which an algorithm was trained, or if it represents something different and novel. While this may seem like mainly a conceptual discussion, it has profound real-world implications for areas such as data privacy, issues of identifiability and confidentiality as well as copyright law. As such, there is currently a lack of nuance in the social science literature on how synthetic data and AI inhabit this tension between frameworks of extraction and ideas of machine generativity.

Objective two: Diversity and difference – To analyse the ways in which synthetic data are generating societal differences as well as a new ethics of diversity. 

Algorithms have been understood as technologies of exclusion and segregation that reinforce stereotypical and culturally entrenched representations in biased outputs based on characteristics such as race, class, and gender. And while many algorithmic systems are often claimed to be ‘blind’ to data attributes such as race and gender, synthetic data constitutes a shift in that they embody an explicit claim to actively generate diverse data points, such as images of black people or text data representing racialised minority populations in a healthcare dataset. The research within objective two focuses on how the algorithmic generation of synthetic data is reconfiguring societal notions of difference and diversity. It examines how synthetic data produce new racial data formations whilst also exploring the wider implications of this reconfiguration for social science notions such as population, bias, and discrimination. In this way, the SYNDATA project will contribute novel ideas and notions to debates around the ethics of algorithms and data, especially in terms of algorithms intersect with questions of difference, diversity, and representation.

Objective three: Bodies and Resistance – To investigate the variegated impacts of synthetic data and algorithms on real bodies and to explore the implications of these on ideas of resistance.

Synthetic data is not simply a proof of concept nor detached from everyday life. Rather, the research will examine the real-world implications of synthetic data on real bodies. While the commercialisation of synthetic data often frames them as ‘risk free’, it is crucial to explore the everyday consequences of training algorithms on synthetic data. This is because synthetic data may nonetheless have a significant impact on people’s lives and bodies – such as how they are rendered recognisable and detectable to the algorithm in new ways. The research will explore how the effects of synthetic data and algorithms are actualised in real human bodies as well as how these effects are unequally distributed. In short, how the ‘synthetic’ is always already ‘real’ in different ways. It is also crucial to better understand what synthetic data means for ideas of resistance and contesting algorithms in society, especially when models trained on synthetic data are claimed to be more ‘ethical’ and ‘fairer.’

Publications

Selected publications

Books

• Jacobsen, B. and Beer, D. (2021). Social Media and The Automated Production of Memory: Classification, Ranking and the Sorting of the Past. Bristol: Bristol University Press. 

Peer-reviewed Articles

• Jacobsen, B. N. AI, Synthetic Data, and the Tensions of the Generative Gap. Accepted and forthcoming in Summer 2026 in special issue at Information Society.
• Beer, D. and Jacobsen, B. (2026). Artificial intelligence, conceptions of distillation, and the reframing of reasoning. Big Data & Society. Online first. 1-12.
• Amoore, L., Bennett, SJ., Campolo, A., Jacobsen, B. N. and Rella, L. (2025). Politics of the prompt: Government in the age of generative AI. Economy and Society. Online first. 1-24.
• Jacobsen, B. N. (2025) Data, algorithms, and the asymmetric politics of extraction. The Sociological Review. Online first, 1-19. 
• Jacobsen, B. N. (2025). Machine Learning, Synthetic Data, and the Politics of Difference. Theory, Culture & Society. 42(3), 41-57.
• Amoore, L., Campolo, A., Jacobsen, B. N. and Rella, L. (2024). A World Model: On the Political Logics of Generative AI. Political Geography, 113, 1-9.
• Jacobsen, B. N. (2024). The Logic of the Synthetic Supplement in Algorithmic Societies. Theory, Culture & Society. 41(4), 41-56. 
• Jacobsen, B. N and Simpson, J. (2023). The tensions of deepfakes. Information, Communication & Society. 1-15.  
• Amoore, L., Campolo, A., Jacobsen, B. N., and Ludovico, R. (2023). Machine Learning, Meaning Making: On Reading Computer Science Texts. Big Data & Society. 1-13.  
• Jacobsen, B. N. (2023). Machine learning and the politics of synthetic data. Big Data & Society. 1-12.   
• Jacobsen, B. N. (2022). ‘You can’t delete a memory’: Managing the Data Past on Social Media in Everyday Life. Sociological Research Online. 27(4): 1003-1019.   
• Jacobsen, B. N. (2022). When is the Right Time to Remember? Social Media Memories, Temporality, and the Kairologic. New Media & Society. 1-17.  
• Jacobsen, B. N. (2021). Regimes of Recognition on Algorithmic Media. New Media & Society. 1-16.  
• Jacobsen, B. N. and Beer, D. (2021). Quantified Nostalgia: Social Media, Metrics and Memory. Social Media + Society. 1-9.
• Jacobsen, B. N. (2020). Algorithms and the narration of past selves. Information, Communication & Society. 1-16.  
• Jacobsen, B. N. (2020). Sculpting digital voids: The politics of forgetting on Facebook. Convergence: The International Journal of Research into New Media Technologies. 27(2): 357-370. 

Teaching

Overview

My teaching mainly focuses on areas related to my research interest, such as the ethics of AI, the politics of memory and social media platforms, the political economy of big tech and data extraction. In addition to teaching on various undergraduate and Masters modules, I also currently convene two undergraduate modules.

Supervision

I welcome new PhD students with an interest in research areas such as memory, social media, algorithms and AI, and data.

Current PhD students include: 

• Paul OrdChurong Geng
• Suemaiya Zaman
• Austin Jones (starting September 2026)
• Sarib Mujtaba (starting September 2026)
• Shutong Tian (starting October 2026)

External Activity

Editorial duties

I am currently an editorial board member on the journals The Sociological Review and Cultural Sociology.