Taming a Buddhist Ghost-AI-Style: Missed Oppurtunities in Re-Generating Stories Beyond the Global North

• In Event: Education policy: actors, ideas and implications

Proposal

The importance of storytelling and story-sharing in educational (research) contexts has been referred to and proven broadly. Recently the importance of collaborative practices has been underlined to give rise to so far ignored, underestimated, or unheard stories and their bearers (Author et al. 2019, Shell-Weiss 2019). Questions related to the documentation of stories and the role of storytellers remain focussed on differences in oral and written traditions, one of the main questions in post-colonial efforts to document stories (Swadener et al. 2008). What happens to storytelling and its essential elements like creating stories for a specific audience, designing the narrative itself, and fulfilling the mediating role of the narrator (Jackson 2013) when a presumably ‘neutral’ AI tells 'our' stories? When we prompt AI to (re-)tell or summarise a central (cultural) narrative? In other words, in comparing digital and ‘traditional’ storytelling modes we ask how AI provides access to different stories and how (far) it changes story-telling. Is it possible to detect biased narrations and how can we challenge biased technologies? In the proposed paper, we combine research on (participatory) storytelling and narratives (Author et al. 2019) with explorative research on the use of AI across (cultural and geographic) borders to scrutinize how narratives and ‘real life’ become entangled (Chubb et al. 2024).
Little has been reported about the supposedly missing story-sharer in the case of AI re-producing stories (Chubb et al. 2024), specifically outside dominating Northern discourses. Moreover, we wonder how diversities within storylines, contents, and ways of storytelling and questioning will be affected by AI-generated stories. And while we assume that participatory (human) storytelling is vital for educational processes, using AI seems to work differently: the artificial storyteller seems to present facts about the storyline in an orderly manner while letting out the more significant cultural framework and missing a culturally appropriate terminology. It is now the role of the audience to detect cracks in the matrix, e.g. the missing context, illogic narrations, or a biased use of terminology. The challenge might now be to follow the story without getting lost in a rabbit hole of new prompts and explanations demanded from or suggested by AI. The knowledgeable interplay between the storyteller and the audience, who are focused on the story or a specific narrative, runs the risk of getting lost in the hypertext of a story (Abbott 2021) that is entertaining but meaningless or in a patchwork of further inquiries that create distractions.
Using the example of Mae Nak, a famous Thai folk tale (e.g. Ancunta 2016, นาวิกมูล n.d.), we explore biases between different language versions and knowledges. The story has been told by many people and inspired many versions and media. Depending on who you ask it is a Ghost, Love, or Buddhist story… Occasions and reasons for telling the story are manifold and include educational contexts. Drawing from data from a storytelling workshop (https://detox.univie.ac.at/en/), different versions of the folk tale (in written, cinematic and oral form), and several online and in-person workshops aimed to question Mae Nak’s story told by AI, we aim to illustrate some of the missing information, omissions and generalizations in the story - not aiming to set the story straight but to underline how fundamental interaction and questions related to making sense of a story remains relevant. Documenting our translingual dialogues and using our exchanges with AI, we will illustrate (missed) opportunities and fears related to flattened stories, how this can become a learning opportunity, and who can become part of new story-sharing collectives or human/communal correctives. When we question the art of AI storytelling and the use of specific terms in the story that we code as biased terms (such as 'exorcism'), we will also ask how AI tries to 'win us over' with 'watery explanations', for example when biased terminology is ‘defended’ with an assumed (white and Northern) audience or when an explanation is declared justified in the next paragraph. Some of the many questions, we encountered along this storied exchange will be guiding our presentation: Does AI lure us further into epistemic injustices, or does it reproduce past ghosts to educate the future culturally literate reader?

References
Abbott, H. Porter (2021). The Cambridge Introduction to Narrative. Third Edition. Cambridge University Press.
Ancuta, K. (2016). That’s the Spirit!. Ghost Movies in Southeast Asia and Beyond: Narratives, Cultural Contexts, Audiences, 306, 123.
Chubb, J., Reed, D., & Cowling, P. (2024). Expert views about missing AI narratives: is there an AI story crisis?. AI & society, 39(3), 1107-1126. https://doi.org/10.1007/s00146-022-01548-2
Jackson, M. (2013). The politics of storytelling: Variations on a theme by Hannah Arendt (Vol. 4). Museum Tusculanum Press.
Author et al. (2019).
นาวิกมูล, เอนก (n.d.). แม่นาก ภาคสมบูรณ์. (Navikmun, A. (n.d.): Mae Nak Classical Ghost of Siam Extensive Version)
Shell-Weiss, M. (2019). The power of narrative: A practical guide to creating decolonial, community-based projects. The essence of academic performance, 1-19.
Swadener, B. B., & Mutua, K. (2008). Deconstructing the global postcolonial. In: Handbook of critical and indigenous methodologies, 31-43.

Authors

• Sabine Krause, Université de Fribourg

• siriparn sriwanyong

• Nawin Ketruam

• Michelle Proyer, University of Luxembourg