Postmodern Elements In “AI DUNGEON”, An Ai-Generated Interactive Fiction

Muhammad, Jagad Lesmana (2026) Postmodern Elements In “AI DUNGEON”, An Ai-Generated Interactive Fiction. S1 thesis, Universitas Andalas.

	Text (Cover and Abstract) Cover and Abstract.pdf - Published Version Download (571kB)
	Text (Chapter 1) Chapter 1.pdf - Published Version Download (434kB)
	Text (Chapter 5) Chapter 5.pdf - Published Version Download (415kB)
	Text (Daftar Pustaka) Daftar Pustaka.pdf - Published Version Download (394kB)
	Text (Full Thesis) Full Thesis.pdf - Published Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

This thesis examines postmodern elements in AI Dungeon, an AI-generated interactive fiction platform, through the lens of Fredric Jameson’s theory of postmodernism and Tzvetan Todorov’s narrative theory of equilibrium. The study aims to identify the Postmodern elements of AI Dungeon in the AI writings, with elements such as fragmentation, intertextuality, crisis of historicity, and the waning of affect, while also looking the way at AI constructs the structure of the narratives. Using two AI-generated stories, Knight and Chosen One, as primary data, the research analyzes narrative progression, emotional expression, and the process of creation dynamics between the player and the AI system. The findings reveal that AI is able to construct human-like narratives and emotions, while also including the instability of postmodernism, as can be seen in how it demonstrates fragmented storylines, a flexible authorship, and non-linear progression. Further in the study, findings of elements of Todorov’s equilibrium model are present; they appear in different tellings, written as a cycle of unresolved forms that reflect the uncertainty of postmodern storytelling. These findings indicate that AI Dungeon does not merely simulate traditional storytelling but transforms classical narrative structures into a hybrid form of digital literature, where meaning is continuously shaped through interaction between human agency and algorithmic processes.

Actions (login required)

View Item