Speaker
Description
Accounting for the expanding relevance of artificial intelligence in academic research, this qualitative study investigates user views of trust in results by ChatGPT within academia. While the American Psychological Association defines interpersonal trust as confidence in another's reliability, trust in AI may involve more complex dimensions including system predictability, reliability, and the rational balance between choice and risk. This research examines these dimensions among 18 computer science master’s students at Bauhaus-Universität Weimar, through semi-structured interviews to understand their experiences and perceptions. Through ad hoc coding and grouping Overarching themes were developed for better understanding of trust. The findings reveal broad views on the reliability of ChatGPT, highlighting the importance of various dimensions including functional utility, Usage Limitations and Trust Management influencing the experience.
