Scholarlink

Text-To-Image (TTI) generators are becoming widely used and are often promoted as “democratizing” the generation of images, but relatively little is known about whether people with different skill sets and literacies use these tools differently, and how their backgrounds influence the quality of the results. This paper investigates the impact of Visual Literacy, Artificial Intelligence (AI) Literacy, and Prompt Engineering Literacy on the use of a human-AI co-creation process for a real-world visual design task. We present a user study (n = 25) examining how people with different literacies interact with Midjourney to complete a visual design task. The results were evaluated by 3 internationally acknowledged visual artists. Our results found no impact of any literacy on the general visual appeal of the generated images, but where participants with high AI literacy reported more understanding of the co-creation process, images created by participants with high visual literacy were rated as better fulfilling the stated visual design task as such. We argue that future work must therefore consider not just the visual appeal of generated images, but also whether they are genuinely useful for a given function.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Algorithmic Creativity: How Visual and AI Literacy Impact the Use of Text-to-Image Tools in Design Tasks

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Alessandro Canossa,
  • Lisa Toender Berger,
  • Lucas Fellner,
  • Willem van der Maden,
  • Jesper Juul,
  • Jichen Zhu

摘要

Text-To-Image (TTI) generators are becoming widely used and are often promoted as “democratizing” the generation of images, but relatively little is known about whether people with different skill sets and literacies use these tools differently, and how their backgrounds influence the quality of the results. This paper investigates the impact of Visual Literacy, Artificial Intelligence (AI) Literacy, and Prompt Engineering Literacy on the use of a human-AI co-creation process for a real-world visual design task. We present a user study (n = 25) examining how people with different literacies interact with Midjourney to complete a visual design task. The results were evaluated by 3 internationally acknowledged visual artists. Our results found no impact of any literacy on the general visual appeal of the generated images, but where participants with high AI literacy reported more understanding of the co-creation process, images created by participants with high visual literacy were rated as better fulfilling the stated visual design task as such. We argue that future work must therefore consider not just the visual appeal of generated images, but also whether they are genuinely useful for a given function.