Pattern #37 · Ideation

AI ideation: sketch your own concepts before you prompt

Otherwise the first render does your thinking.

Track Aai-ideationdesign-fixationcreativitydesign-processgenerative-ai

Do

Sketch alone first. Cluster the wall. Then prompt for the gaps: "Give me five concepts unlike these."

Don't

Open the workshop with the model's first answer. Enjoy your nine Build-a-Breeze Castles.

The rule. Put your own concepts on the wall before you ask a model for any, then prompt for the directions the wall is missing.

Why. Design fixation is "blind adherence to a set of ideas or concepts limiting the output of conceptual design" (Jansson & Smith, 1991, as cited in Wadinambiarachchi et al., 2024). Generative tools feed it. In a CHI 2024 experiment, 60 people sketched chatbot avatars after seeing one example; those using Midjourney fixated more on it and "produced fewer ideas, with less variety and lower originality" than a no-support baseline (Wadinambiarachchi et al., 2024). Across a team, the effect compounds: in five experiments, ChatGPT-assisted brainstorming produced narrower idea sets, with the drop in diversity significant in 37 of 45 comparisons (Meincke et al., 2025; Needleman, 2025). In creative writing, AI ideas made individual stories more creative and the stories more similar to each other (Doshi & Hauser, 2024).

None of these studies tested sequencing directly. Sketching first is the practice that follows from them: the team's own concepts become the baseline, and the model is asked to extend the set instead of setting it.

Seen in the wild. FigJam AI can work on ideas people already wrote: it sorts a selection of stickies by theme and summarizes them (Figma, n.d.).

References

  1. 01

    Wadinambiarachchi, S., Kelly, R. M., Pareek, S., Zhou, Q., & Velloso, E. (2024). The effects of generative AI on design fixation and divergent thinking. In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems (CHI '24). Association for Computing Machinery. https://doi.org/10.1145/3613904.3642919

    https://arxiv.org/abs/2403.11164
  2. 02

    Meincke, L., Nave, G., & Terwiesch, C. (2025). ChatGPT decreases idea diversity in brainstorming. Nature Human Behaviour, 9(6), 1107-1109. https://doi.org/10.1038/s41562-025-02173-x

    https://www.nature.com/articles/s41562-025-02173-x
  3. 03

    Needleman, E. (2025, May 14). New in Nature: ChatGPT decreases idea diversity in brainstorming. Mack Institute for Innovation Management, The Wharton School. https://mackinstitute.wharton.upenn.edu/2025/new-in-nature-chatgpt-decreases-idea-diversity-in-brainstorming/

    https://mackinstitute.wharton.upenn.edu/2025/new-in-nature-chatgpt-decreases-idea-diversity-in-brainstorming/
  4. 04

    Doshi, A. R., & Hauser, O. P. (2024). Generative AI enhances individual creativity but reduces the collective diversity of novel content. Science Advances, 10(28), Article eadn5290. https://doi.org/10.1126/sciadv.adn5290

    https://doi.org/10.1126/sciadv.adn5290
  5. 05

    Figma. (n.d.). Use AI tools in FigJam. Figma Learn. Retrieved October 9, 2026, from https://help.figma.com/hc/en-us/articles/16822138920343

    https://help.figma.com/hc/en-us/articles/16822138920343