Artificiell intelligens och dess påverkan på mjukvaruutvecklares arbetsprocesser: En kvalitativ studie
2024 (Swedish)Independent thesis Basic level (degree of Bachelor), 10 credits / 15 HE credits
Student thesisAlternative title
Artificial Intelligence and its consequences on the work processes of software developers : A qualitative study (English)
Abstract [en]
This study explores how software development processes are affected by the integration of AI tools in the workplace. The aim was to clarify the changes that have occurred and the consequences of the adoption of AI into software developing work processes, and how skills and workflow therefor might have changed.The structure of this study was made through qualitative semi-structured interviews with 10 professional software developers in different fields and with various experience about their use of AI-tools, specifically ChatGPT and Github CoPilot, supplemented with a literature review of academic sources about what processes a software developer work can consists of, and sources containing earlier research on the consequences AI technology have on both students and professionals in the field of software development.The result of this study shows that there are several work processes that have been changed or automated with the use of AI tools, and the software developers who took part in empirical collection all thought AI tools have helped in a more efficient work process, where more of their usual work gets done quicker and easier than before. Specifically in processes such as documentation, coding, problem solving, repetitive tasks and learning new concepts. However, there are still aspect of software development tasks that AI tools cannot yet replace or change which includes tasks that involves comprehending larger concepts or understanding and communicating with customers to ensure requirements of a system are met and properly integrated in the start and the end of a project. The study also show that AI tools can halt workflow by for example introducing bugs or unnecessary code in its suggestions, and therefore the importance for a competent professional to review any of these.The results of this study also show that respondents have become somewhat dependent on AI tools but are still cautious with it. However, most respondents agree that if forced, they still hold the ability to go back to the workflow they had before the use of AI tools.
Place, publisher, year, edition, pages
2024. , p. 52
Keywords [en]
Artificial intelligence, Software development, AI tools, Github Copilot, ChatGPT, Processes
National Category
Information Systems, Social aspects
Identifiers
URN: urn:nbn:se:hv:diva-21804Local ID: EXI500OAI: oai:DiVA.org:hv-21804DiVA, id: diva2:1871367
Subject / course
Informatics
Educational program
Systemutveckling - IT och samhälle
Supervisors
Examiners
2024-06-272024-06-172025-09-30Bibliographically approved