Prompting Mindsets

Abstract

The study explores the concept of eliciting mindsets over the generative artificial intelligence (AI) within the framework of how Gen-Z students perceive artificial intelligence tools across different situations that are context-dependent. Learning about the topic of user behaviour and the specific prompting of behaviour, in particular, has emerged as an important subject of study as generative AI technologies, such as ChatGPT, increasingly become part of everyday academic, professional, and social activities. The study is based on a qualitative and exploratory approach and involves semi-structured and in-depth interviews with Gen-Z people who are active users of generative AI technologies. The fact that the study is based on the real-life illustrations of prompting is a crucial element that makes it possible to obtain the actual patterns of behaviour as well as the introspective observations. The theme analysis is applied in the study to analyse how users develop, modify and refine prompts based on various contextual factors, such as audience expectations, evaluation criteria, time limitations and perceived consequences. The findings reveal that prompting is a dynamic process that is cognitively motivated and is based on situational demands instead of a fixed or technically precise operation. Participants will show a clear change in the various prompting mindsets, which include exploratory, execution-oriented, creative and accuracy-driven, based on whether they are working in academic (marks-oriented) or social/creative (engagement-oriented) situations. It is also shown that prompting behaviour is reflective and iterative and functions as a metacognitive loop whereby the user refines his/her inputs until he/she gets the desired results. Another important aspect of the study to note is the significance of the expected audience on the choice of prompting techniques, where it is revealed that users dynamically adjust their vocabulary, tone and structure based on who would evaluate or utilise the product. This presents prompting as a social practice as opposed to a solitary technological engagement. Altogether, by introducing prompting as an integrated behaviour, both cognitively and socially, this research contributes to the growing body of knowledge about generative AI."