The librarian, Andrew Gray, has discovered a surprising trend in scientific studies published last year. He found that the use of certain words like meticulously, intricate, commendable, and meticulous had suddenly increased in their English versions. After analyzing five million studies, Gray concluded that tens of thousands of researchers may be using ChatGPT or similar AI language generation programs to write or enhance their research papers.
Examples of this trend are evident in various studies. Some authors have used text generated by AI programs in their introductions, conclusions, or even in their responses to reviewers. Gray estimates that at least 60,000 scientific studies in 2023 were written with the help of ChatGPT or similar tools. While some researchers use AI appropriately to enhance their work, others may be relying too heavily on these programs without verifying the accuracy of the results.
AI language models tend to disproportionately use words with positive connotations, which influences the way scientific studies are written and reviewed. Studies have shown that reviewers who use AI in their evaluations tend to repeat positive words more frequently, even in prestigious journals. Gray and other researchers are concerned about the impact of this trend on the quality and authenticity of scientific research.
The sudden increase in the use of terms like intricate, meticulous, and commendable in scientific studies has raised questions about the reliance on AI in academic writing. Gray’s analysis reveals a significant shift in the language and tone of scientific studies, which may have long-term implications for the field. He warns against the potential consequences of this trend, including the generation of low-quality or insubstantial research.
Researchers like Ángel María Delgado Vázquez and Jeremy Nguyen note that non-native English speakers often use AI tools to improve their language skills and writing. While AI can be a valuable resource for language translation and refinement, its widespread use in academic writing raises concerns about the authenticity and originality of research. As the scientific community grapples with the impact of AI on research, transparency and verification are essential to maintaining the integrity of academic work.