Understanding Adoption Dynamics: GenAI Assistants in Software Development

Abstract

Prior research has shown the huge potential of using GenAI assistance in various domains, including Software Development (SD). However, theses studies are mainly focused on the capabilities of these tools rather than the adoption dynamics. Especially research in the SD domain is missing. Therefore, this study will investigate which factors influence the adoption of GenAI assistants in SD. To that end, I conducted semi-structured interviews and a Creative Problem Solving (CPS) session with experts within the Software Development Lifecycle (SDLC). The results show that the adoption of GenAI assistants is currently lacking. These factors that influence the adoption were retrieved by means of a thematic analysis of the results. The results show that the lack of personalization and missing human capabilities are critical factors influencing the adoption, resulting in a lack of trust in these GenAI assistants. Additionally, there seems to be a social stigma on the usage of GenAI assistance in SD. The CPS session revealed that integration of nuanced answers, the creation of persona’s and profiles, and the lead-by-example from higher level employees turned out to be important factors in overcoming these adoption challenges in future GenAI assistant implementations.