The following topics will be addressed:
- What are the effects of detailed prompting (e.g., briefing on, e.g., purpose, participants, participant organizations, desired outcomes) as opposed to just providing the brainstorm question and other ideas coming in?
- Does the AI react to the contributions of other (human) participants? If so, how?
- In an anonymous brainstorming, how do contributions of the AI blend in with human contributions?
- Are there thresholds of acceptability for human participants regarding the quantity (number and length) of AI contributions?
- Are stage instructions to the AI regarding the mood (e.g., be humorous), perspective (e.g., take a marketing perspective) or desired effect (e.g., encourage divergence) effective in shaping outcomes?
The research is supported by XLeap GmbH. XLeap will provide a version of its facilitation software that includes a prototype of the Brainstorm workspace integrating ChatGPT model 4 with configurable prompts. The master student will have direct access to the company’s CTO and the senior developer of the project."