Skewed Artificial Intelligence: Flagging Embedded Cultural Practices in Children's Stories Featuring “Alice and Sparkle”

Abstract

AbstractWe live in an age of artificial intelligence (AI). Open AI, an American artificial intelligence organization, claims that generative chatbots are here to stay. Although computer technology remains unevenly distributed, the presence of AI continues to be felt in many areas of daily life. Further, AI development happens in few countries whose cultural practices and beliefs tend to be embedded in AI output and functionality. Western and Global North perspectives tend to dominate conversations about AI. How might AI technology function equitably in an ethno‐racially and linguistically diverse world? This paper explores intersections of technology, coloniality, and bias to contextualize the user‐experience for individuals in groups characterized as high, medium, and low‐resourced languages. For whom does AI function in our globalizing world? Engaging with this question through a languacultural (Agar, 1994) lens, I employ a language and culture perspective to guide inquiry into AI’s output and impact, based on a children’s storybook. Specifically, I deploy Kenyan, Indonesian, and Taiwanese languacultural perspectives of three participants in a culturally cognizant reading (Washington, 2023) of a product of chatbot‐human composing, ‘Alice and Sparkle’, a children’s book illustrated by MidJourney and co‐authored by ChatGPT and Ammaar Reshi. Overall, through scrutiny of encoded cultural tropes and values inscribed within text and images, this paper proposes considerations for literacy pedagogy and research that stem from a pursuit of inclusive AI.