AbstractBuilding on the risks of digital extractivism, this talk presents a case study on a traditional Sámi herding practice as a stepping stone for more responsible AI design. By attempting to translate herding strategies into formal logic via Social Practice Theory, we perform a failure analysis to identify where digitization fractures. The results demonstrate that while logic can coordinate the syntax of herding, by mapping roles and routines, it fails to capture the semantic depth of embodied skills and historical memory. We argue that these epistemic gaps constitute protective boundaries that define where the machine’s knowledge ends and sovereign human wisdom must begin.
SpeakersNellie Pierre is a MSE student in Interaction Technology and Design at Umeå University. She focuses her research on holistic perspectives, questioning what developing technology in ways that benefit humanity, the environment and society really means from different perspectives. Currently, she is writing her master’s thesis on the role of transparency and interpretability in AI-systems related to epistemic justice.
Tuva Falk holds an MSc in Interaction Technology and Design from Umeå University. Her work critically examines the intersection of technology, society, and ethics, with a specific focus on the limits of machine logic. Driven by a passion for responsible AI, she investigates how digitization impacts cultural practices and embodied knowledge. Looking ahead, her ambition is to pursue further research exploring how we can design technology that safeguards human wisdom and respects ethical boundaries.
Please note. The subtitle is generated by AI and may contain some errors.
https://www.umu.se/en/events/fraiday-navigating-the-epistemic-limits-of-ai_12165694/