Nurse Researchers' Experiences and Perceptions of Generative AI: Qualitative Semistructured Interview Study.

Background: With the rapid development and iteration of generative artificial intelligence, the growing popularity of such groundbreaking tools among nurse researchers, represented by ChatGPT (OpenAI), is receiving passionate debate and intrigue. Although there has been qualitative research on generative artificial intelligence in other fields, little is known about the experiences and perceptions of nurse researchers; this study seeks to report on the topic. Objective: This study aimed to describe the experiences and perceptions of generative artificial intelligence among Chinese nurse researchers, as well as provide a reference for the application of generative artificial intelligence in nursing research in the future. Methods: Semistructured interviews were used to collect data in this qualitative study. Researchers mainly conducted interviews on the cognition, experience, and future expectations of nurse researchers regarding the use of generative artificial intelligence. Twenty-seven nurse researchers were included in the study. Through purposive sampling and snowball sampling, there were 7 nursing faculty researchers, 10 nursing graduate students, and 10 clinical nurse researchers. Data were analyzed using inductive content analysis. Results: Five themes and 12 subthemes were categorized from 27 original interview documents as follows: (1) diverse reflections on human-machine symbiosis, which includes the interplay between substitution and assistance, researchers shaping the potential of generative artificial intelligence, and acceptance of generative artificial intelligence with alacrity; (2) multiple factors of the usage experience, including individual characteristics and various usage scenarios; (3) research paradigm reshaping in the infancy stage, which involves full-process groundbreaking assistive tools and emergence of new research paths; (4) application risks of generative artificial intelligence, including intrinsic limitations of generative artificial intelligence and academic integrity and medical ethics; and (5) the co-improvement of technology and literacy, which concerns reinforcement needs for generative artificial intelligence literacy, development of nursing research generative artificial intelligence and urgent need for artificial intelligence-generated content detection tools. In this context, the first 4 themes form the rocket of the human-machine symbiosis journey. Only when humans fully leverage the advantages of machines (generative artificial intelligence) and overcome their shortcomings can this human-machine symbiosis journey reach the correct future direction (fifth theme). Conclusions: This study explored the experiences and perceptions of nurse researchers interacting with generative artificial intelligence, which was a "symbiotic journey" full of twists and turns, and provides a reference and basis for achieving harmonious coexistence between nurse researchers and generative artificial intelligence in the future. Nurse researchers, policy makers, and application developers can use the conclusions of this study to further promote the application of generative artificial intelligence in nursing research, policy making, and product development.