Incongruent Priorities in Software Development Teams Using AI

Implementing AI tools is an investment of resources in all software development teams. Members of software development teams have adopted AI tools differently based off of different priorities in programming. This interview study engages software developers (N = 11) in interviews about their software development practices since larger scale AI deployment in their respective companies. Our findings show four separate clusters of software developers coming out of these AI augmented teams with top priorities ranging from code quality, teamwork, up-skilling/learning, and speed of development. These different clusters were developed into personas which can support organizations in understanding how, why, and when AI tools are helpful to their software teams. We engaged both junior and senior developers in the interviews to understand if the clusters were experience dependent and found a spread of expertise across all categories sans the Democratizing Deployer which had only highly experienced members.