Research on English Creative Writing Based on Artificial Intelligence – Taking Deepseek as an Example

Abstract

In the contemporary era, artificial intelligence has progressively permeated all facets of human existence. Its impact on the educational sector, in particular, is profound. Creative writing, as a conventional pedagogical approach in English instruction, is frequently employed to enhance students’ linguistic competence and literary cultivation. Consequently, investigating the supportive role of artificial intelligence in English creative writing assumes significant importance. This study, based on the continuation exercises of endings for Langston Hughes’ Early Autumn authored by both students and artificial intelligence, conducts a comparative evaluation of their respective merits and demerits across three dimensions: narrative design, diction and syntax, and descriptive techniques. The findings indicate that AI-generated texts demonstrate superior narrative construction, more precise diction, and a capacity to accurately emulate the original author’s style through meticulous language use. In contrast, student compositions exhibit greater strengths in creativity and authenticity, often unconstrained by the source text and achieving effects that are “unexpected yet reasonable.” Therefore, in future English pedagogy, artificial intelligence should be integrated into students' creative writing processes to optimize their expressive and rhetorical outcomes, while ensuring that students’ autonomous creativity is not stifled.