Reproduce “From Prompt Engineering to Prompt Craft” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Joseph Lindley, Roger Whitham
Researchers from Lancaster University propose reframing human-AI interaction from "prompt engineering" to "prompt craft," emphasizing an iterative, material-like engagement with generative AI models. Their practice-based Design Research, spanning three projects, develops interaction design strategies and demonstrates how tangible and embodied inputs can enable more intuitive and accessible control over diffusion-based image generation.
Key insights
- Generative AI models exhibit a metaphorical "materiality" with "soft edges" that can be understood and manipulated experientially through embodied interaction.
- Reframing interaction from "prompt engineering" to "prompt craft" fosters a more intuitive, iterative, and design-led approach to working with AI's inherent uncertainty and complexity.
- Careful constraint and 'holistic control' of the AI's vast possibility space, alongside tailored interaction design strategies, are essential for creating meaningful and accessible user experiences.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.