Affiliated Projects

This dissertation examines how distributed forms of anti-war mobilization in Russia, emerging in the aftermath of the full-scale invasion of Ukraine, transform into a mode of collective agency termed survival organizing. Grounded in the case of the Russian anti-war ecology, the research investigates how diverse and self-organized actions initially emerged without central coordination, often in exile or under conditions of extreme repression. These actions were marked by a rejection of traditional hierarchies and the legacy of the liberal opposition, instead embracing principles of horizontality, autonomy, care, and ethical responsibility.

The work traces how these distributed initiatives—ranging from mutual aid networks and digital solidarity campaigns to care work and feminist anti-war interventions—form a dense and dynamic field of resistance. What connects them is not a shared ideology or centralized leadership, but a shared commitment to acting in the present while prefiguring alternative futures. These practices are often fragile, small-scale, and embedded in everyday life, yet they hold transformative potential precisely because they build infrastructures of support and relational continuity in times of collapse.

Through this lens, the dissertation develops the concept of survival organizing—a mode of collective action that moves beyond reactive survival to become a proactive, world-building strategy. It is “survival with a plus sign”: a form of organizing that sustains life and enables political subjectivation under conditions of profound precarity. Survival organizing draws on the logic of autopoiesis, not only maintaining existing structures but generating new forms of relationality and resistance that are open-ended, non-linear, and often anonymous. Importantly, this form of agency is decentered; it is not located in individual actors or organizations but embedded in practices, relations, and infrastructures that can be taken up and actualized by others—even those yet to come.

By focusing on the distributed, prefigurative, and ethical dimensions of anti-war organizing, this dissertation offers a rethinking of collective action under authoritarianism and sociatal inertia. 

Machine learning (ML) is a key area of artificial intelligence (AI) research which receives respective investment and attention but which also reignites familiar debates about the societal impact of AI. In light of issues arising with ML-based technologies – such as algorithmic discrimination  – many demand a democratization of AI, calling for a participatory approach and the inclusion of more diverse communities into the creation of AI systems as well as for an establishment of public AI infrastructures and accessibility for everyone. Researchers particularly emphasize the need for a counterbalance to the power position which big technology companies inhabit within the AI industry. By positioning themselves at the center of application, infrastructure and discourse, these companies succeed to steer the AI debate for their own profit, shaping not only what direction its research and development takes, but also the way AI is regulated. 

It is also these companies, as I present in this PhD project, that have appropriated the democratization discourse to their economic benefit. They are promoting their products to be beneficial and empowering to all of its users, allowing for societal progress; to be inclusive and representative of everyone; and to enable collaborative efforts in AI development. In my PhD project I critically analyze the AI democratization discourse as propelled by big technology companies such as Google, IBM and OpenAI by detailing the narratives they establish but also the corporate strategies they pursue within this discursive frame. Here, I propose not to merely center the debate of democratization around notions of access. Rather, as I argue, it is important to go beyond viewing AI as powerful and opaque entity which needs to be democratized and to shed light on specific ML techniques and the power dynamics in which they emerge. Taking up this perspective means shifting the focus from considering power in the form of ownership of AI infrastructures to a more detailed consideration of how power operates through ML algorithms. In this way, we can trace the dominance big tech companies exert in detail, by analyzing how they shape AI values and infrastructures according to economic rationales.