Return to search

AI- assistenter - en självklarhet i svenska organisationer? : En kvalitativ studie om medarbetares upplevelser avseende möjligheter och utmaningar med användningen av stora språkmodeller i det dagliga arbetet.

This study aimed to provide insights into the opportunities and challenges perceived by employees regarding the integration of large language models (LLMs) into their daily work routines, focusing on the Microsoft Copilot tool. Despite existing research showcasing the potential benefits of LLM-driven AI tools such as efficiency and fostering creativity, there remained a gap in understanding how such tools, particularly Copilot, influenced employees' work routines and tasks. Employing a qualitative approach, this study conducted 10 semi-structured interviews to investigate employees' experiences and perceptions. Hermeneutics served as the guiding methodological framework, allowing for a nuanced exploration of participants' experiences. The findings revealed a multifaceted use of Copilot within the IT company, ranging from idea generation to support in creation processes. Copilot facilitated both formal and informal learning by providing quick access to internal information and feedback, thereby streamlining tasks and freeing up time for other activities. However, challenges such as the limited Swedish language support and formal onboarding hindered its full utilization. Additionally, employees' attitudes toward technology and AI significantly influenced their adoption of Copilot, with technologically inclined individuals being more inclined to embrace and explore its functionalities. Furthermore, the study highlighted changes in work routines, including altered thought processes, support for idea generation, and verification of generated information.  Information security emerged as a notable concern, necessitating caution to prevent leakage of sensitive data. Discrepancies were observed between how employees presented Copilot to clients and their actual utilization, suggesting areas for improvement such as tailored training programs. In conclusion, while Copilot showed promise in enhancing work processes and facilitating learning and idea generation, there were challenges to address, particularly in terms of mindset shifts and integration into existing workflows. Nevertheless, being in the early stages of adaptation there was room for improvement and opportunities to enhance usage, ultimately benefiting the organization's operations.

Identiferoai:union.ndltd.org:UPSALLA1/oai:DiVA.org:umu-227054
Date January 2024
CreatorsKlaréus, Amanda, Enkvist, Alma
PublisherUmeå universitet, Pedagogiska institutionen
Source SetsDiVA Archive at Upsalla University
LanguageSwedish
Detected LanguageEnglish
TypeStudent thesis, info:eu-repo/semantics/bachelorThesis, text
Formatapplication/pdf
Rightsinfo:eu-repo/semantics/openAccess

Page generated in 0.0021 seconds