Reproduce “Exploring the Potential of Metacognitive Support Agents for Human-AI Co-Creation” 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

Frederic Gmeiner, Kaitao Luo, Ye Wang, Kenneth Holstein, Nikolas Martelaro

This research explores how voice-based metacognitive support agents can enhance human-AI co-creation in design, particularly within 3D CAD generative design workflows. The study demonstrates that such agents, whether employing Socratic questioning or structured planning, significantly improve design outcomes and foster deeper cognitive engagement among users, addressing common challenges like intent formulation and problem exploration.

Key insights

  • Metacognitive support agents can effectively counteract cognitive offloading, leading to more thorough problem exploration and precise intent formulation in GenAI design workflows.
  • Socratic questioning specifically fosters critical evaluation, self-checking, and mental simulation, helping designers identify and correct design flaws.
  • Structured planning activities (like Free Body Diagrams) and proactive suggestions enhance design process efficiency, aid in catching errors, and maintain user focus.

Summary by alphaXiv.

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