Hopeful Failure: How Collaborative Design Fiction Reimagines AI
Published in Conditionally Accepted
Authors: Jeffrey Basoah, Katharina Reinecke, Daniela Rosner, Ihudiya Finda Ogbonnaya-Ogburu
While the number of people using AI is growing, the number of people making core AI decisions remains limited. Advocates call for opening up the development of algorithmic systems to a wider range of perspectives, interests, and methods, with particular attention to racial exclusions and harms. This paper responds to this suggestion with two design fiction workshops where 10 Black American participants imagine futures with and against AI. We introduce Exquisite Tellings, selectively reading in-progress stories while co-developing design fiction plots. Across both workshops, participants repeatedly imagined moments of technological failure, including algorithmic breakdowns and mechanical malfunctions. Rather than signaling collapse, these failures surfaced forms of resourcefulness, enabling characters to reconnect with personal and collective capacities obscured by automation. We argue that analyzing specific instances of `hopeful failure’—where challenges in AI development reveal broader social possibilities—can help scholars and critics better understand the emerging effects of AI on society.