A Framework for Collaborating a Large Language Model Tool in Brainstorming for Triggering Creative Thoughts

Among various techniques developed to foster creative thinking, brainstorming is widely used. With recent advancements in Large Language Models (LLMs), tools like ChatGPT have significantly impacted various fields by using prompts to facilitate complex tasks. While current research primarily focuses on generating accurate responses, there is a need to explore how prompt engineering can enhance creativity, particularly in brainstorming. Therefore, this study addresses this gap by proposing a framework called GPS, which employs goals, prompts, and strategies to guide designers to systematically work with an LLM tool for improving the creativity of ideas generated during brainstorming. Additionally, we adapted the Torrance Tests of Creative Thinking (TTCT) for measuring the creativity of the ideas generated by AI. Our framework, tested through a design example and a case study, demonstrates its effectiveness in stimulating creativity and its seamless LLM tool integration into design practices.

process that includes becoming sensitive to problems, deficiencies, information shortages, missing elements, identifying challenges, searching for solutions, making estimates and hypotheses, modifying these hypotheses in response to the identified shortcomings, and trying one of these solutions

how to use prompt engineering to generate creative ideas is neglected.

suggested by creativity research, usefulness and novelty are two essential aspects of creativity

Drawing inspiration from previous studies on the essential elements of design approaches and frameworks, we have developed the Goal-Prompt-Strategy (GPS) framework, which integrates prompt engineering for the utilization of LLM) tools. The purpose of this framework is to help develop effective prompts when using Generative AI for brainstorming purposes, particularly during the design process.

Strategies play a pivotal role in our framework as established in prior research for guiding design thinking. However, within our framework, strategies serve a different purpose. Instead of directing the designer’s thinking, they shape how an LLM tool generates responses. We recognize that an LLM tool operates based on interactions between questions asked by the users and answers generated by an LLM tool. Research on thinking skills (e.g., creativity, decision-making, problem-solving, and critical thinking) highlights the importance of questioning as a method for fostering human creativity and solution generation (Wayne Allison et al., 1986; Penick, 1996; National Advisory Committee on Creativity and Cultural Education, 1999; Raz et al., 2023). We believe that various question-asking strategies or techniques may not only influence LLM responses but also the flow of the conversation. Therefore, strategies in the context of question-and-answer interactions are used to form proper and purposeful questions that guide LLM’s thinking. By treating an LLM tool as a collaborative partner, strategies introduce various ways of questioning for stimulating innovative ideas and solutions; that is, they influence how initial prompts may be adjusted or refined to elicit more effective responses. The following lists nine strategies that are essential for creative prompt engineering.

four dimensions are defined as follows: (a) fluency: the ability to develop large numbers of ideas; (b) flexibility: the ability to produce ideas in numerous categories; (c) originality: the ability to produce unusual or unique ideas; (d) elaboration – the ability to adapt abstract ideas into realistic implementations