How ChatGPT shapes a new reality of writing: Is there a place for humans in an artificial world?

Abstract

Aim. This study explores the impact of Artificial Intelligence (AI) -driven text generation on authorship recognition and its implications for education. Specifically, it examines how individuals with varying levels of experience using ChatGPT perceive AI-generated texts and whether editing affects their ability to distinguish between human and AI-authored content. The study aims to inform strategies for integrating AI literacy into education while addressing challenges in authorship attribution. Methodology. A quantitative study was conducted with 85 participants, categorized based on their ChatGPT usage: no experience (n=22), occasional use (n=31), and daily use (n=32). Participants evaluated texts across five stylistic genres (literary, journalistic, scientific, philosophical, and promotional) and attempted to determine authorship (human, AI-generated, or AI-edited). The data were analyzed using statistical methods, including the Kruskal-Wallis test, ANOVA, and Pearson’s correlation analysis. Results. The findings indicate that participants struggled to differentiate between human and AI-generated texts, with an average accuracy of 5.48 out of 15. The ability to identify authorship varied by genre, with philosophical texts being the easiest to recognize and journalistic texts posing the greatest challenge. Experience with ChatGPT did not significantly improve recognition accuracy. Editing AI-generated texts further blurred the distinction between human and machine authorship. Conclusions. The results highlight the need for educational approaches that enhance critical literacy and awareness of AI-generated content. AI tools can serve as collaborative writing aids, but their integration into learning requires ethical considerations and adapted assessment strategies. Future research should expand the sample size and investigate the effects of long-term AI exposure on writing perception.

Ähnliche Arbeiten