Digital Ojar: Whose knowledge feeds the machine?

A new project by Dr Maya Indira Ganesh, Associate Director and Co-director of the Narratives and Justice programme at CFI, asks what it means to be marginalised in the world of the AI. Digital Ojar is a digital representation of Quilt Ojar, a physical quilt made by the Arctic Indigenous Visual Artists Network (AIVAN). The quilt’s 21 tiles, handmade with materials sourced from Indigenous homelands across the Arctic, tell stories of human and nonhuman ecologies of the region. Visitors to the Digital Ojar website can click on each tile to learn about the artist and the story behind it. “Ojar” in the Saami language is a word for the ripples made by a stone thrown into water. The project addresses the question: what happens to marginalised stories, languages, and cultural knowledge when they enter the internet and AI’s infrastructures, and who decides? “The same tools that can help save endangered languages and traditional knowledge can and do copy, strip, and reuse that knowledge without asking permission or giving anything back,” says Dr Ganesh. “Data scraped to train AI, cultural archives digitised by outside institutions, platforms built for growth rather than trust all raise the same question: who is this really for?” These are broader questions of community and cultural resilience that converge with the politics of the internet that AI ethics and data policy will have to confront. In the Relations section of the website, is a series of new tiles created by the artist Kira Xonorika using open-source AI image generators; they are her interpretations of the AIVAN community’s traditional motifs. They connect the site to a broader network of Indigenous data and AI projects. Kira “My intention was to bring energy and honour into a project that touches artificial intelligence, and to do so with transparency. These are digital images, AI-generated, but deeply guided by human intention, research, and care.” The process of making these tiles was slow and unfolded in conversation with the AIVAN community, explaining how their data will be used. Digital Ojar offers a provocation to consider both ambivalence and agency in adopting AI, to consider the multiplicity of relationships and impacts in its use and proliferation. Digital Ojar is part of the Imagine Technoscience Virtual Arts Residency, supported by the Yale University-MacMillan Fund awarded to Kalindi Vora (Yale University), Nishant Shah (Chinese University of Hong Kong), and Maya Indira Ganesh (University of Cambridge). At the centre of the project is a simple idea: communities should have a say in how their own knowledge is collected, stored, used, and shown – including when it ends up feeding an AI system trained on scraped or donated material. Dr. Tatiana Degai and Kira Xonorika were the artists supported by the Cambridge node, and Arunabh Pal Singh worked on the website design and building. Professor Kalindi Vora commented: “Imagine Technoscience came together as a way to look across several continents (North America, Europe and Asia) to ask: what kind of imagination and knowledge-making specific to art practice do we need in the face of the proliferation of generative AI? The goal was to discover new approaches to imagine, speculate, model, and narrate the not-yet-imagined futures of AI technologies to inspire us to see possibilities of affordances and challenges they offer. We looked to artist practitioners for a virtual residency who were developing new language, lenses, and analytical frameworks to understand the ways that art practice can establish new relationships to the future that depend on these technologies, but reimagined.” More info: Digital Ojar Main image credit: Anna Sakmarkina, digitally re-worked by Kira Xoronika.