Idea Co-Construction or Friction: Collaboration Patterns in Human-AI Co-Writing (Poster 18)

• In Event: AERA Poster Session Thursday 3:35 p.m.
  In Poster Session: (Generative) Artificial Intelligence in Education

Abstract

The advances of Generative Artificial Intelligence (GenAI), including Large Language Models that can co-create content with humans, offer great promises for augmenting learning and creativity. We design a LLM co-writing tool that provides writing suggestions and supports students to regulate the writing process. The tool’s output encourages regulatory strategies such as planning writing tasks, monitoring progress, and reflecting on the evolving stories. We collect audio and video-recorded interviews with graduate students, who co-wrote stories about future technology designs with the GenAI tool. In this paper, we present three cases that showcase different collaboration patterns with the AI co-writer—representing idea co-construction and friction. We discuss implications for learning technology design to support active idea negotiation between humans and AI.

Authors

• Saríah López-Fierro, Utah State University

• Ha Nguyen, University of North Carolina - Chapel Hill