Exploring ChatGPT's impact on student-team ideation outcomes for new product development: a pilot study

Ideation techniques such as associative-thinking methods are commonly used to explore design proposals. However, limited experiences and knowledge in young designers can constrain diverse and meaningful design solutions. Emerging artificial-intelligence technologies, like ChatGPT, provide easy access to a global knowledge base which could inform associative-thinking outcomes. ChatGPT excels at generating lists of user-specified topics to accelerate learning with access to decades of gathered online experiences and insights. This study hypothesised that using ChatGPT to inform associative-thinking techniques would improve student idea generation compared to analogue methods in a new product development workshop. Product ideas were represented on Post-it notes, and outcomes were measured by fluency, flexibility, and originality. Thirty-five undergraduate students (first-year freshmen to fourth-year seniors) from Brigham Young University participated in two innovation workshops. One utilised ChatGPT in team ideation efforts, and the other used analogue methods. Over 75 percent of students had engineering related majors of study while less than 25 percent were non-engineering disciplines. All students were equally taught associative thinking techniques, and the ChatGPT group had additional training on software usage. Results show that fluency and flexibility outcomes were slightly lower in the ChatGPT group. In originality, the analogue group averaged twice the ideas of the ChatGPT group. Self-reported performance of flexibility and originality were lower in the ChatGPT group, but higher for fluency. Ideation effectiveness, enjoyment, and empowerment were all lower in the ChatGPT group. Observations revealed that ChatGPT-assisted teams had increased team interactions. Future research might benefit from longer ideation sessions and visualisation training.