Personalized media and its poor compromise between anonymity and identification

We are delighted to invite you to a seminar organized jointly by DALOSS and the Digital Culture Research Cluster, featuring a presentation by Randi Heinrichs from Leuphana University Lüneburg.

Abstract

The digital economy markets convenient, personalized services while preserving user privacy. Their personalized services promise to understand and predict what you are and want without being able to identify you. As my presentation will discuss, this paradoxical alignment of identification and anonymity is facilitated by the application of a recognition principle that targets user types. In other words, contrary to the promise, personalization is not about a unique person and their individual needs, but about creating a stand-in that is assumed to be like that person and similar others. To illustrate that ideal or fake types have nevertheless real consequences for real people, I compare in the attached dissertation chapter “Configurations of identity and identification – the particular in the general ” current recognition systems with earlier identification methods of the late 19th century, like Sir Francis Galton’s composite portraits as well as Alphonse Bertillon’s Bertillonage. What has changed in digital cultures are the methods used to design, capture and cluster segments or ‘types’ of people. Based on my ethnographic fieldwork in the digital economy, including a 10-month participant observation in a German tech company (especially a team of developers that created a local social media platform about neighbourhoods) and interviews with various stakeholders of the dominating Tech Hub of Silicon Valley ranging from digital activists and privacy advocates, UX-designers and data scientist, I will discuss how different methods to create those stand-ins turn users into ‘known unknowns’. More specifically, my talk addresses the creation of user personas as a key tool of design thinking as well as the more data-driven approach behind recommendation algorithms called data neighbouring. Building on the emerging field of critical data studies, I discuss the fact that the omission of particular features, in other words of identifying data, is not only not necessary for the creation of those user segments, but is even a hindrance to emphasizing the supposedly common, similar characteristics. From this point of view, anonymization and the systematic deletion of data ironically no longer appears as an antidote to personalisation, but as necessary. Anonymity becomes a feature of what I call ‘personalization’ and is thus entangled in the same political inequities that these recognition systems re-produce. This understanding of anonymity goes beyond concerns of individual privacy but opens up a broader debate about the current logic of non-identification and their hegemonic systems of subjectification, exclusion and discrimination.

Sign-up

Please RSVP to nannab@hum.ku.dk by 11 March 2024. Refreshments will be provided, and a chapter of Randi Heinrichs' dissertation will be shared.

About the Speaker

Randi Heinrichs recently defended her PhD thesis on "Configurations of identity and identification" at Leuphana University's Centre for Digital Culture. Her research focuses on anonymity, social media platforms, and data neighbourhoods. She is now embarking on a postdoc position in a research project on smart cities at Leuphana. Randi's interests lie at the intersection of digital cultures, technology, and issues surrounding anonymity, security, fairness, and discrimination. She has been associated with esteemed research groups at UC Berkeley and Simon Fraser University and is actively involved in editorial roles within academic circles.