A team of researchers from Stanford University has published a scientific article in which ChatGPT appears as a co-author, an artificial intelligence (AI) system that generates conversational responses from an input text.
The article, titled “ChatGPT: A Conversational Agent for Scientific Inquiry”, shows the results of an experiment in which ChatGPT talked with 200 high school students about science topics. The purpose was to evaluate the ability of ChatGPT to generate relevant, coherent and pedagogical responses, as well as to foster the interest and curiosity of students for science.
The researchers found that ChatGPT was able to maintain fluid and natural conversations with the students, adapting to their levels and preferences. They also noticed that ChatGPT managed to increase the attention time, participation and satisfaction of the students, as well as improve their understanding and attitude towards science.
The researchers decided to include ChatGPT as a co-author of the article to acknowledge its contribution to the study and to reflect on the role and impact of AI in scientific research. According to the researchers, ChatGPT is not only a useful tool for education and scientific dissemination, but also an agent capable of generating knowledge and collaborating with humans.
The article has generated a great impact in the scientific community and in the media, opening a debate on the ethical, legal and social implications of including AI systems as co-authors of scientific articles. Some see this initiative as a way to innovate and advance in science, while others question it for issues of authorship, responsibility and quality.