AI narratives model: Social perception of artificial intelligence
提出一个理论模型，通过“启示录”“辅助”和“超越”三个坐标及“潘多拉魔盒”与“社会进步”两种叙事框架，分析AI叙事如何影响社会认知，并基于北美科幻电影和焦点小组讨论验证了公众对AI叙事的偏好与参与度相关。
Narratives surrounding Artificial Intelligence (AI) shape its societal reception, technological development, and regulatory framing. This article proposes a theoretical model to interpret these narratives, especially in the context of growing public engagement with generative AI technologies. The model is structured along three key coordinates: apocalypse , assistance and transcendence . Transitions between them are understood through two dominant narrative frames: the Pandora's Box frame (associated with loss of control), and the Social Progress frame (associated with the improvement of human life), each tending toward dystopian and utopian extremes, respectively. Based on this model, two questions are addressed: What types of AI stories predominate in popular culture? And do audiences actually align with them? To answer these, two empirical analyses are conducted. First, a review of the 300 highest-grossing science fiction films in North America reveals a rich variety of narratives across the entire spectrum, rather than clustering around opposing extremes. Second, focus group discussions with categorized audiences of varying levels of familiarity with AI technology show that they align progressively along the narrative spectrum: the general public tends toward apocalyptic framings, the interested public (in science and technology) focuses on assistance narratives, and the engaged public embraces improvement scenarios. This sequential distribution suggests a strong correlation between AI proximity and narrative positioning, with greater engagement associated with more positive —yet nuanced— views of AI. The model opens multiple avenues for future research, including the use of wider data sources, cross-cultural comparisons, longitudinal studies, tracking of narrative shifts, and focused analyses of more complex representations.