On the afternoon of October 21, the 13th session of the Qizhen Qiushi Media Forum took place at Zhongxi Bookstore – Dazhong Bookstore (Zhejiang University Campus Bookstore), hosted by the Digital Communication Research Center of Zhejiang University, the College of Media and International Culture, and Zhejiang University's Digital Social Sciences Convergence Research Initiative. The lecture featured Professor Claudia Aradau from the Department of International Politics at King’s College London and Professor Tobias Blanke from the University of Amsterdam, specializing in Artificial Intelligence and the Humanities. They presented their work, Algorithmic Reason: The New Government of Self and Other. The event was chaired by Professor Huang Guangsheng, a researcher at Zhejiang University under the “Hundred Talents Program,” with dozens of faculty and students in attendance for an engaging exchange.
Professor Claudia Aradau focuses on the role of technology in shaping (in)security, particularly how digital technologies redefine security and surveillance practices while influencing the interplay between security, democracy, and critique. Professor Tobias Blanke, with expertise in computer science and political philosophy, is a faculty member at the University of Amsterdam's Institute for Logic, Language, and Computation and a visiting professor of Social and Cultural Informatics at King’s College London. His research interests include applying AI and big data tools in the humanities.
During the lecture, Professors Aradau and Blanke introduced the background of their book Algorithmic Reason, explaining the development and application of the concept of algorithmic reason. The lecture was divided into three parts: the background of algorithmic reason, its core elements, and the mechanisms of algorithmic governance.
Part 1: Background of Algorithmic Reason
Professor Aradau began with a case study involving humanitarian organizations using algorithms to assist decision-making, which highlighted the growing role of algorithms in contemporary society. She then introduced the book’s “scene” methodology, where each chapter opens with a controversial scenario to analyze how different participants problematize the practices, knowledge, and implications of algorithmic reason. Scenarios are viewed as windows into the world and stages for conflict. This methodology frames disputes as a specific mode of public contestation, different from the binary structure of conflict. Instead, it adopts Cyril Lemieux’s triadic structure of contestation, involving not only the disputing parties but also the public audience of the disputes.
Aradau noted the increasing attention on artificial intelligence and big data in recent years. Social and political researchers often highlight the revolutionary epistemological and political ruptures brought about by algorithms, while critical scholars in the humanities and social sciences warn that overemphasizing algorithms' disruptive nature risks obscuring persistent social and political inequalities. Aradau advocated for a middle ground to understand the social transformations algorithms bring. The concept of algorithmic reason provides a lens to trace the interplay of continuity and discontinuity in governance practices and to examine how algorithms function across various domains.
Algorithmic reason, she explained, represents a cognitive framework that helps us understand how digital technologies influence social governance, creating divisions and hierarchies. Algorithmic governance is not merely a technical issue but a complex phenomenon involving societal, racial, and power dynamics. At its core, algorithmic governance determines what constitutes (ab)normal behavior and prioritizes interventions for addressing anomalies.
Part 2: Core Elements of Algorithmic Reason
Professor Blanke elaborated on two key components of algorithmic reason.
Blanke discussed how algorithmic reason reshapes relationships between individuals and groups, speech and action, and self and other. The book’s first chapter uses the Cambridge Analytica scandal as a starting point to illustrate how algorithms decompose and reassemble small and large-scale entities, forming a political rationality for governing individuals and groups. This logic redefines the boundary between speech and action, establishing a “truth-doing” model central to algorithmic reason.
In the second chapter, Blanke analyzed controversial scenarios of predictive policing, focusing on how CivicScape’s algorithms operate through subtle variations in workflows and data. Each element can influence the overall outcome. Thus, the second core element of algorithmic reason involves partitioning abstract computational spaces or “feature spaces” within machine learning.
Part 3: Interventions in Algorithmic Governance
Professor Aradau explored mechanisms for intervening in algorithmic governance. In the book's final chapters, she examined the shift from being governed by algorithms to making algorithms governable. This analysis focused on how algorithms blur boundaries between domestic and international spheres, state power, and the global power of major tech companies.
Today, tech companies play an increasingly significant role in algorithmic governance. While states attempt to regulate these companies through legal measures, they also collaborate with them in co-governing the digital world. For example, social media platforms shape public opinion and political attitudes by algorithmically curating user content. In this context, the “authors” of algorithmic governance extend beyond state institutions to include diverse stakeholders such as private enterprises and technology developers. Governance thus involves complex interactions not only between states and individuals but also between states and corporations.
Algorithmic ethics has become a focal point of governance, often dominated by technical experts who exclude marginalized groups most affected by algorithms, such as racial and gender minorities. Aradau and Blanke illustrated this exclusion with two frictional scenarios: the Google employees’ petition against military AI projects and collaborative hacking efforts. Drawing on Bonnie Honig’s concept of “public things,” they argued that public things are dynamic entities formed through collective action and emotional connections.
Public interventions, such as petitions, serve as “little tools of friction” that challenge algorithmic governance. Although the Google employees’ petition ultimately failed to halt the company’s military collaboration, these tools sparked broader discussions on algorithms' role, achieving accountability and democratizing the ethical discourse traditionally dominated by technical experts.
The lecture concluded with an engaging discussion between attendees and the speakers. Topics included changes in human-computer interaction after the advent of ChatGPT, differences in algorithmic governance mechanisms across political and social contexts, privacy concerns in facial recognition, and strategies for promoting algorithmic ethics through public participation.