Generative artificial intelligence and decision-making: evidence from a participant observation with latent entrepreneurs

Purpose – This paper investigates how individuals without prior entrepreneurial experience interact with generative artificial intelligence (GenAI) when making decisions related to an entrepreneurial task. The study aims to develop a conceptual framework to underline the main factors influencing human–GenAI interaction, clarifying how GenAI can support, rather than substitute, human decision-making in complex contexts. Design/methodology/approach – The authors collected data using a participant observation methodology by employing a conversational GenAI tool. Additionally, a post-observation survey was administered to gather participants' opinions about the technology. Secondary data were also collected. The authors analysed the data through an inductive thematic analysis. Findings – The analysis led the authors to develop three main themes: (1) Human–machine interaction: Taking or Giving? Participants could interact with the machine by adopting two different behaviours, taking and giving knowledge or opinion. (2) Humans as leaders of the conversation, humans gained responses which did not always meet their expectations, deciding to lead the conversation. (3) Perspective on GenAI: Operational tool vs Learning source, participants leveraged the machine's operational and informational capability or intended it as a learning source. Originality/value – This paper provides several insights for understanding how individuals without previous experience interact with GenAI when making decisions in a complex setting. This paper provides theoretical and practical contributions supporting the Human + paradigm.