Prompting visions: Modern participation in science

The “Prompting visions“ app serves as the starting point for an interactive touchpoint at conferences, trade shows, events or workshops. »Prompting visions« not only generates diverse and personalized images but also provides valuable data for scientific analysis.

Multidirectional engagement

Science communication and the involvement of society in research processes are essential for establishing innovations collaboratively. Science communication in Germany is often based on written and spoken words. In particular, when scientists wish to systematically document societal dialogue as feedback for their own research field, the question arises as to how this can be captured and utilized. Researchers often rely on short surveys conducted during events for this purpose. However, these traditional surveys rarely capture the emotional depth and individual attitudes necessary for in-depth, participatory research. The use of technical jargon further complicates mutual understanding among experts from science, politics, business and society.

Experience from past research projects shows that participatory formats can foster multidirectional exchange – but often fail due to linguistic, cultural or educational barriers, and that direct engagement with complex topics can lead to uncertainty, feeling overwhelmed and negative emotions.

The rapid development of generative artificial intelligence and the associated image-generating diffusion models are creating new possibilities: people without creative training are now able to express content through their own visualizations. For science communication, this development holds considerable potential, as for the first time, participants in exchange and participatory formats can convey their thoughts visually to the outside world. Personal opinions, emotions and visions on topics outside one’s field of expertise can be generated as images with the help of AI – without any drawing talent whatsoever.

The challenge lies in the fact that the visual formulation of abstract content requires design expertise and relevant prompt engineering knowledge.

The idea: “Prompting visions” modernizes participatory science and uses image-generating AI as a translation tool between stakeholders. A customizable web application with a user-centered interface design collects individual data on a predefined topic via a short survey. The responses are automatically embedded into image-generating AI models, allowing the personal AI image to be generated in real time.

How “prompting visions” works:

1.  In preparation, questions and answers (in the form of sliders, free text or multiple-choice) are selected from the “prompting visions” template or created independently by the researchers 
2. On-site, users simply answer the shared survey via their smartphone.
3. The answers provided are compiled in the background into an optimized AI prompt, and an AI image is generated within seconds.
4. The resulting personalized image can be saved as a digital file or printed as a high-quality postcard.

The collected data and generated images can serve as valuable insights in technology-oriented research projects to better understand public interest, impact and expectations. The method is based on psychological research findings and current approaches in science communication.

The added value of “prompting visions” lies in the following areas:

Insights gained:
1. The keywords entered can be further utilized for analysis. The generated image also encourages self-reflection, and the multitude of different visions provides new inspiration for research projects.

2. Strengthening participation:
Thanks to easy access via smartphone and AI elements, the tool offers a playful introduction to new or familiar topics. Everyone is empowered to contribute their ideas and opinions, regardless of their level of knowledge.

3. Promoting communication
The tool also serves as a helpful icebreaker, as the AI-generated images can spark unconventional conversations, particularly among people who do not know each other.

The combination of data collection and image-generating AI makes “Prompting visions” an attractive tool for innovative group and event formats. Abstract topics become tangible, and interdisciplinary exchange is encouraged. The intuitive use of the web application allows complex research topics such as climate change, the circular economy, or the energy transition to be visualized and made accessible on an individual level.

• Low-threshold survey tool
  The structure of the survey and the app reduces barriers to entry. This form of data collection is a more engaging method for both researchers and respondents compared to standard questionnaires.

• Introduction to topics for groups or events
  The live-generated images serve as an incentive and starting point for open discussions and interpretations. Participants’ spontaneous responses enable agile work and quick reactions. The app and the instant photo printing station have been successfully tested multiple times at events.

“Prompting visions” thus contributes to science with social responsibility and can be easily integrated into future projects to bring together diverse levels of knowledge and interests.