Towards human-AI-collaboration in brainstorming: Empirical insights into the perception of working with a generative AI. (EDU800 Week #12 Annotated Bib)


Memmert and Tavanapour (2023) completed a study to look at the experience of participants in using generative AI to support brainstorming. While brainstorming in a group setting is commonly found to be a positive experience for participants, this study focused on two products: cognitive stimulation and ‘free-riding’. This mixed-method study looked at 24 participants using Chat GPT-3 as a support for brainstorming. Participants were tasked with navigating a web-based application designed to provide AI suggestions to the prompts provided by the participants. The authors collected qualitative evidence through a survey post-study in addition to the log data from the application used by the participants. The results of the study lend credence to the two components focused on. Results implied there is a gain in process through cognitive stimulation with the AI system. However, the authors acknowledge the opportunity for ‘free-riding’. To combat this potential, the authors suggest “careful interaction design” to prevent human to AI system interactions from being received and used without consideration of potential errors.

The use of the qualitative measure as a means to collect data was appropriate and allows for context and perception from the participants to be flushed out. The survey included 18 open-ended questions which were designed to address the two components addressed within the study. I also found the procedure to be particularly insightful within the liberty that was provided to the participants using AI. There were minimal instructions provided to the participants outside of including at least one AI suggestion. This freedom allowed the authors to freely evaluate the true human behavior among the participants. This also accurately reflects the brainstorming process which can often be ever-changing and non-linear. The choice to reflect the procedure this way was a strength of the overall study. The study also addresses that much of the future research needs to focus on how to quantify the components of cognitive stimulation and free-riding, major areas where most educators view AI.

Within my current role as an educator, AI has become a part of my life and my career. Many of my colleagues share concerns and perspectives regarding the use of AI among students. I am interested in the prospects of future research looking at free-riding and the cognitive benefits of AI, especially depending on the age group. However, as someone with exposure to AI previously, I feel less impacted by this article. It spoke to me as something relevant to my career and aspects of what I see with my colleagues. That said, the findings of this study feel inconsequential as they are primarily common knowledge among educators who actively use AI. In regards to my research, I would like to expand on the use of AI to brainstorm and the effects it has academically and in motivation/engagement. Specifically in inquiry, what are the impacts on students when using AI as a means to brainstorm product creation, etc.

Sources
Memmert, L., & Tavanapour, N. (2023). Towards human-AI-collaboration in brainstorming: Empirical insights into the perception of working with a generative AI.

Leave a comment