Reproduce “Interaction Design with Generative AI: An Empirical Study of Emerging Strategies Across the Four Phases of Design” 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

Marie Muehlhaus, Jürgen Steimle

An empirical study by Saarland University researchers systematically explored how Generative AI tools (GPT-4, DALL-E3) can support interaction design across all four phases of the user-centered design cycle. The research documented effective usage strategies and prompting patterns, assessing AI-generated artifact quality and revealing both versatile support and critical challenges in practical application.

Key insights

  • Generative AI, when integrated through an "iterative AI-in-the-loop meta-design cycle," effectively augments human creativity by rapidly generating diverse ideas and providing critical feedback.
  • Effective prompting strategies are adaptive, with single structured prompts excelling in early phases, iterative prompting beneficial for complex technical tasks, and persona patterns valuable throughout.
  • GenAI's role is complementary, requiring human oversight for quality assessment, functional realism, bias mitigation, and addressing ethical concerns to ensure practical and responsible design outcomes.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.