Paper short abstract:

This paper presents the analysis of dialogues between scientists and the non-expert public on X/Twitter about generative AI. We identified scientists engaging with the public, the discussion topics and genres, and the roles scientists and the public play in creating highly engaged dialogues.

Paper long abstract:

As dialogues between scientists and the public are emphasized as vital components of science communication (Einsiedel, 2008; Irwin, 2021), we examined generative Artificial Intelligence (AI) as a topic of such dialogue. Generative AI began attracting the public’s attention when an easy-to-use planform, ChatGPT, became widely available in November 2022. The impact of AI, more specifically generative AI, on society has been touted as enormous (Sabherwal & Grover, 2024), yet there are many unknowns about this innovative technology. Thus, it is natural for the non-expert public to seek out expertise from scientists. Social media platforms allow the non-expert public to learn from and engage with scientists about diverse topics online. We analyzed dialogues between scientists and the non-expert public on X, formerly Twitter, about generative AI during the early stage of ChatGPT’s public release. We used topic modeling, content analysis, and regression analysis to identify who the scientists are and to communicate about generative AI, what and how scientists discuss generative AI, what roles scientists and the non-expert public play while engaging in dialogue about generative AI, and what roles scientists and the non-expert public play to create high engagement in an online platform. Our findings shed light on dialogues between scientists and the non-expert public about an emerging technology using Kling (1996)’s genre regarding discourses about computerization as well as online knowledge sharing roles (Hara & Sanfilippo, 1996) as frameworks to analyze the content of posts generated by both scientists and the non-expert public.