The Psychological Effects of AI-Assisted Programming on Students and Professionals

Artificial intelligence (AI) tools have become integral to the coding workflow, facilitating productivity through auto-completion, code suggestions, and chat dialogues. More than ever, the psychological relationship between software engineers and AI partners requires further exploration, mirroring the need to understand the psychological aspects of pilot-navigator roles these tools partly simulate, as highlighted in earlier studies. [Goal] The presented research aims to investigate the programmer behavior change and psychological effects of AI-assisted programming on professionals and undergraduates. [Methods] The authors performed a series of seven experimental programming sessions on the undergraduate student sample, subjecting them to programming in solo, pair, and AI-assisted settings. Following the experimental sessions, five semi-structured interviews were conducted with the experimental participants from the academic realm and another five with experienced users of AI programming tools from the professional realm. The interviews were analyzed using thematic analysis in an essentialist way, allowing for a comprehensive understanding of the participants’ experiences, feelings, and attitudes toward AI-assisted programming. [Results] A total of ten themes were identified, half shared across the two realms, with 51 constituent codes selected from the professional and 35 from the student interview transcripts. The realm-agnostic themes were: "Effectance", "Intrinsic Motivation", "Perceived AI Personality", "Dynamics of Human and AI Pairing," and "Paradigm Shifts". The realm-specific themes included: "Personal Growth and Development", "Prospects", "Ethical Considerations", "Safety", and "Effects on Learning Processes". [Conclusion] The presented research illuminates the profound potential of using AI as a programming partner in simulated pilot-navigator roles in both professional and academic realms. It underscores the importance of understanding its effects on psychological well-being and human experience. Incorporating the revealed psychological aspects of human-AI interaction, its benefits, limitations, and dynamics will be pivotal for successful AI implementation and continued evolution in software engineering contexts.