Browsing by Author "Berg, Totti"
Now showing 1 - 2 of 2
- Results Per Page
- Sort Options
- Älykellot lääketieteen sovellutuksissa
Insinööritieteiden korkeakoulu | Bachelor's thesis(2021-12-05) Berg, Totti - The effects of generative artificial intelligence on ideation quality
Insinööritieteiden korkeakoulu | Master's thesis(2024-08-19) Berg, TottiThis thesis examines the impact of generative artificial intelligence on ideation quality, particularly comparing AI-generated ideas (e.g., by ChatGPT-4) with those from company professionals during group brainstorming sessions. The work examines both abstract and technical topics, and explores how the presence of AI influences human ideation capabilities across different themes. The research method employed group brainstorming sessions conducted in two formats: traditional and AI-assisted. Participants, who were members of the company’s research and innovation team, generated ideas in both setups, and these ideas were evaluated based on workability, relevancy, and specificity. The evaluation was carried out by the company’s Innovation Management team. The results indicated that ChatGPT-4 produced slightly higher-quality ideas on average than humans for abstract topics. However, for technical topics, company experts generated better ideas than the AI. When collaborating with AI, the average quality of human ideation declined for technical topics but remained the same for abstract topics. Specifically, the AI-generated ideas scored higher in workability and relevancy of abstract topics, but human-generated ideas excelled in terms of workability and relevancy on technical topics. Notably, AI’s ideas scored significantly higher in specificity, resulting in more evaluatable ideas compared to human-generated ones, which scored lower in specificity and were more commonly not evaluated. In collaborative settings, AI also helped humans generate more specific ideas. In summary, generative AI has the potential to enhance ideation in certain contexts, but its benefits vary depending on the topic. Additionally, the impact of AI on human ideation is complex and requires further research to determine optimal collaboration methods.