Artificial intelligence chatbots (AICs) are advanced systems capable of generating and processing human-like text, and are being increasingly integrated in various fields, including education. Despite their potential to significantly impact learning, little is known about university students’ and postdoctoral fellows’ (US&PD) views on AICs in educational settings. This study investigated the familiarity, perceptions, and factors influencing adoption of AICs by US&PDs in the life and medical sciences. We conducted a cross-sectional online survey. Recruitment involved two approaches: (1) using R script on PubMeD metadata to extract contact details of corresponding authors with recent MEDLINE-indexed publications, and (2) collecting publicly listed contact information of program administrators from the top 50 global, English-speaking universities, as ranked by the Quacquarelli Symonds (QS) University World Rankings. Both authors and administrators were contacted and requested to forward the survey to US&PDs. The survey was administered via SurveyMonkey from February 2 to March 18, 2024, with two reminder emails sent between February 14 and 26, 2024. A total of 1209 responses from 73 countries were analyzed. Most respondents identified as female (62.07%) and were enrolled in doctoral (40.48%) or master’s programs (17.55%). Over 63% were familiar with AICs, with ChatGPT being the most used (60.3%). While many recognized the educational value of AICs, concerns about reliability and integration into academia persisted. Calls for more training and institutional support were common. The study underscores the potential and challenges of AICs in education. While enthusiasm exists, significant concerns remain about their implementation, requiring targeted training and policy development.
Ng et al. (Sat,) studied this question.