By Dominika Lisy (Linköping University)

My contribution to the conference is one from the margins where bodies, affective experience, and relational boundaries can be found. From a feminist new materialist perspective, materiality and its agency are key aspects, which do not get much attention in understanding emergent technologies, such as AI chatbots or social robots. In my PhD research, I studied the emergence of social robots, which challenges the relationality between human and non-human in specific ways because robots are purposefully designed and researched with and through certain ideas about human form, behaviour, and cognition. Techno-optimist or techno-pessimist stances that influence robot development circle around very rigid ideas about the human, which range from going beyond human limitations to protecting human uniqueness, respectively. The foregrounding in feminist new materialist perspectives of entanglement (Barad, 2007) and the embrace of a posthuman vulnerability (Shildrick, 2024; Daigle, 2023) are important antidotes to such rigid understandings of the human. Feminist new materialist tools (e.g., compound words, diffraction, and situated knowledges) have thus far made it possible to highlight the ambiguity and unknowability of things that hitherto seemed clear-cut. This has been crucial for material agency and its role in nurturing a posthuman ethics. However, with the development of robot bodies, there is a risk of losing critical force in contesting irresponsible design choices and research agendas, in which robot bodies end up imposing themselves on human lives and experience. Therefore, in my research, I explored how the encounter with robots changes boundaries of the human body and experience. I developed my concept of in/flexible boundaries from reading together Grosz’s (1994) concept of the ‘Möbius strip’ and the concept of the ‘agential cut’ by Barad (2007). I then used this understanding of boundaries to analyse my experience with robots. In particular, in my encounter with touch robots, I found that where the robot starts and where my body ends shifts constantly. I argue that it is crucial to pay attention to these boundary shifts and that integrating knowledge and experience about one’s body into the research of technology is paramount. Only through the investigation of boundaries as material-discursive points of contact is there a chance to develop a radical responsibility (Shildrick, 2024) and feminist sensitivity for the human/non-human relationality.

By Ninell Oldenburg (University of Copenhagen)

Philosophy of technology has long engaged in questions about how technical systems shape knowledge, human being, and social relations. This talk argues that feminist epistemologies, as developed within this tradition, offer an undervalued lens for holistically analyzing, critiquing, and reimagining the ecosystem and technology of AI. Specifically, current debates in the philosophy of AI follow a largely Western, analytic tradition, in which timely topics like AI consciousness and welfare, Silicon Valley ideologies, and risks of AI are predominantly looked at through the isolated stances of the philosophy of cognitive science, politics, ethics, and epistemology. Feminist epistemology combines these into a richer, encompassing critique.

The talk begins with a brief introduction of some of the main feminist epistemological critiques, such as the critique of the view from nowhere (Haraway), hyperrationality, disembodiment, and hierarchical dualist conceptions of science (Hayles, Lloyd, Plumwood), and the hierarchy of knowers (Adam).

First, the view from nowhere critique posits that every knowledge is embodied and situated and that attempts to universalize, generalize, and abstract such situated knowledge necessarily hide the positionality of the knowledge producer. This is especially dangerous if this one occupies a privileged social position. Second, the critique of hyperrationality and hierarchical dualisms shows how in the history and in contemporary science and philosophy, the male was upheld as abstract, rational, and disembodied, whereas the female was neglected as concrete, emotional, and embodied. Feminist epistemology illuminates what knowledge gaps this has produced and how to fill them. Third, the hierarchy of knowers is the ordering of people whose knowledge counts as valid, due to the specific type of produced knowledge and/or their social position, and those who are systematically excluded from the knowledge production process. In the realm of AI, this raises questions about whose voices shape research agendas, whose experiences count as data, and how those excluded from the ecosystem might instead become active participants in it.

The second half of the talk then uses the three topics mentioned above to briefly introduce how the current discourse approaches them, and then delineates what feminist epistemology can add concretely. In short, traditional critiques offer separate accounts for ideologies, ethics, or epistemologies. Feminist epistemology combines these three aspects to offer a rich, encompassing critique of this interesting and complex net of actors and stakeholders.

More specifically, AI consciousness and welfare currently lack an account for the underlying dominating assumptions of computational functionalism as a disembodied, male one. This raises important questions about funding priorities with an epistemologically biased imagination of the world, but also ontological questions about the nature of AI. Silicon Valley ideologies have been critiqued as anti-feminist, fascist, capitalist, transhumanist, amongst other things, but lack a coherent critique of the interplay thereof. Lastly, AI risks, specifically, existential risk research, has been critiqued on the grounds of its excluding sociology, sealed off political economy, rigid ethics, and, ofcourse, narrow epistemology. However, it lacks a holistic account for the specific problem that sits at the intersection of all of these subdisciplines.

By Anja Salzmann and Paul Prinsloo (University of Bergen)

In his latest book, "Against progress" Slavoj Žižek (2025) relates the story of a magician who performs a trick with a canary which disappears when its cage is flattened, only for the canary to reappear, miraculously, in the hands of the magician. After the show the magician removes the dead canary from the flattened cage and throws it in the trash where many other dead canaries lie. The 'magic' was impossible without a dead bird and concealing what has been sacrificed was essential for sustaining the 'magic'. For Žižek, this illustrates the dialectical notion of progress - for each new claim of 'progress' there must be a concealed squashed bird.

Central to our critique of AI as 'Progress' is Žižek’s claim that in a linear, universal and universalising, teleological narrative of progress, its power and effectiveness conceal the residue of what has been sacrificed. Contrary to a teleological perspective Žižek suggests a radical and recursive redefinition of the concept of progress entailing at every step of the way, becoming aware of "the squashed bird that was the victim of this progress" as basis for redefining progress. This is the essence of an ethics of/for squashed birds.

This paper maps the contours of a critique of the notion of AI as/and progress as a basis for reconfiguring an ethics of AI in the context of education. The "religion of progress" pronounced from "the pulpits of progress" with its "prefabricated futures", its "intensifying state of emergency" and imaginaries as "futurecraft", are based on a specific understanding of progress as linear, inevitable, its foreclosure a "civilizational necessity" and a "redemption." The linguistic, structural and political framing of AI as 'Progress' and as inevitable is not only a "category error" but it also "suppresses alternatives, narrows acceptable debate, and obscures trade-offs that would otherwise remain visible." In this understanding of AI as/and Progress, AI is not just a tool for realising this vision, but a way of seeing the world where everything becomes (an exploitable) resource(s) [Bestand] to be used, stored or replaced. This 'enframing' [Gestell] pushes us in "a particular onto-historical direction, subtly but pervasively shifting humanity’s basic understanding of what is and what matters" and who we are.

Educational policy and practice are not immune to the algorithmic logic of these AI magicians with their machinations of magic promising faster, more efficient and more effective personalised learning - all in the name of Progress. Education is seen as broken and in need of technological magic.

An ethics of/for squashed birds transcends AI's enframing of all things as resources [Bestand] or "expendable pieces", "standing-by to be quantified, ordered, and optimised with maximal efficiency." It is, in its very essence, anticipatory, linking designs and decisions to consequences, aligning incentives with responsibility, and distributing accountability in service of the living birds in cages awaiting their turn in the theatre of magic late-capitalism.