Telling Creative Stories Using Generative Visual Aids
Project Overview
The document explores the role of generative AI in education, particularly its impact on enhancing creativity in storytelling through the use of visual aids. It describes a study in which writers were provided with AI-generated visuals that corresponded to their writing prompts. The findings revealed that these writers produced stories that were not only more creative and original but also more complete and visually engaging, leading to greater enjoyment in the writing process. However, the study also highlighted a potential drawback: the visuals sometimes hindered the integration of the initial prompts, indicating a trade-off between divergent thinking (creativity) and convergent thinking (focusing on the prompt). Overall, the document illustrates how generative AI can serve as a powerful tool in educational settings, fostering creativity while also presenting challenges that educators must navigate.
Key Applications
Visual Stories, a web interface that aids story writing by providing generative visuals based on prompts.
Context: The educational context involves adult participants engaged in creative storytelling, specifically writers who are fluent in English.
Implementation: Participants were divided into two groups: one with access to generative visuals (V+) and one without (V-). They wrote stories based on given prompts and used a web interface to toggle through visuals.
Outcomes: Writers in the V+ group found the process more enjoyable, produced longer and more creative stories, and reported that the visuals inspired their narratives.
Challenges: The visuals may distract writers from effectively integrating the prompts, leading to decreased convergent thinking and potentially over-reliance on visual aids.
Implementation Barriers
Cognitive Barrier
The use of visual aids may distract writers from focusing on the textual prompt, leading to poorer integration of the prompt in their stories. Additionally, writers may develop a dependency on visual aids, which could hinder their verbal creativity when such aids are not available.
Proposed Solutions: Future implementations could dynamically update visuals based on the story's progression to keep the writers engaged with the textual prompts. Encouraging writers to develop their own creative processes without visual aids in some exercises could help mitigate this risk.
Project Team
Safinah Ali
Researcher
Devi Parikh
Researcher
Contact Information
For information about the paper, please contact the authors.
Authors: Safinah Ali, Devi Parikh
Source Publication: View Original PaperLink opens in a new window
Project Contact: Dr. Jianhua Yang
LLM Model Version: gpt-4o-mini-2024-07-18
Analysis Provider: Openai