Abstract This study examined the ability of English-major college students to distinguish between AI-generated and human-authored texts based on aesthetic qualities, addressing gaps in cross-culturals, empirical, and collaborative research on detection. Previous studies, often limited by short texts, Western focus, and sparse empirical data, struggled to identify nuanced aesthetic differences. This study distinguishes individual (G1) and collaborative (G2) aesthetic assessments to test whether social deliberation enhances the detection of AI-generated narratives, with implications for collaborative literary pedagogy. Group deliberation was hypothesized to mitigate individual cognitive biases in aesthetic judgment. Using a mixed-methods design, 79 students (40 individuals, 39 in thirteen groups of three students) evaluated four long-form texts presented in Chinese (two human-authored, two AI-authored; over 1500 characters each, equivalent to approximately 1000 words in English) using the Simplified Literary Analysis Framework. Results showed no significant aesthetic differences, with individuals rating texts higher than groups did, indicating that collaborative scrutiny amplified critical evaluation. Qualitative data revealed groups emphasized AI’s modernized language and abrupt structures, while individuals noted stylistic similarity but overlooked cultural depth. The study bridged gaps by employing extended texts, cross-cultural Chinese contexts, and aesthetics-trained participants, confirming AI’s stylistic mimicry but cultural limitations. These findings suggest that collaborative analysis enhances detection by focusing on cultural nuances and offering practical strategies for identifying AI-generated texts. For education, integrating AI texts into curricula through group exercises can sharpen students’ aesthetic discernment, preparing them for AI-influenced literary landscapes and advancing both theoretical understanding and practical teaching approaches.
Guan et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: