https://doi.org/10.36719/2706-6185/44/193-200
Abderrahim Cheikh
Kasdi Merbah University Ouargla
https://orcid.org/0009-0008-0353-539X
cheikh.abderrahim@univ-ouargla.dz
Comparative Analysis of AI-Generated Verses: Imitating Edgar
Allan Poe’s Literary Style
Abstract
Literature and creative writing are two areas where artificial intelligence (AI) is opening new avenues and is questioning the entire idea of authorship and artistry. This work demonstrates how generative AI, and more specifically ChatGPT, can mimic the stylistic features of Edgar Allan Poe, using The Raven as a reference. Two experimental methodologies are employed: (1) a structured atmospheric generation with pre-written verses to establish thematic and linguistic consistency, and (2) a dialogue-based approach in which an AI writing assistant interacts with a melancholic narrator to co-create poetry dynamically. The AI-generated texts are evaluated against Poe’s original work using a structured framework that assesses theme and atmosphere, language and diction, rhythm and meter, and narrative imagery. The results indicate that AI can effectively capture the melancholic tone, gothic aesthetic, and rhythmic complexity of Poe’s poetry, though minor deviations in archaic diction and strict metric consistency persist. The study highlights the strengths of AI-assisted literary creation while acknowledging its limitations in fully replicating human intuition and creative spontaneity. These findings contribute to the broader discourse on AI’s role in literature, demonstrating its potential as both an analytical tool and a creative collaborator in reimagining classical literary styles.
Keywords: Generative AI, creative writing, The Raven, literary emulation, artificial intelligence, poetry generation, authorship