Artificial intelligence, especially generative AI, is increasingly complementing various processes in society, including the field of user experience design. Due to the gap between research and practical application of AI-tools, this master thesis explores how they can support the user-centered design process on a practical level. The research focuses on evaluating the usability of these tools in different phases of the design process, using the redesign of the Slovenian Alpine Association (Planinska zveza Slovenije) website as a case study. The goal was to examine in which stages AI can effectively assist designers and what its potential and limitations are in this context.
The theoretical part analyzes previous research and AI-tools applicable to various phases of the design process. The experimental part involves the use of tools such as ChatGPT-4o, Attention Insight, Relume, Equally AI, DALL-E 3, and others. These tools were used to analyze survey data, generate heatmaps, conduct automated accessibility evaluations, and create user personas, information architecture, wireframes, and graphics for the final design solution.
The results show that artificial intelligence can effectively support user-centered design, especially in the phases of ideation, prototyping, and market research. It also proves useful for fast and cost-effective testing of intermediate design iterations. However, the research concludes that AI cannot replace human judgment and that final testing with real users remains essential. AI should be seen as a valuable support tool for co-creation and process optimization, not as a means for full automation.
|
