Abstract
This session examines the principles of effective AI collaboration through my experience rebuilding a WordPress codebase with Claude over nine months, during which I witnessed the tool evolve from an often-frustrating, sometimes-useful novice to a pro-level, agentic collaborator. I’ll demonstrate the role of expertise in AI collaboration by showing real conversations where my domain knowledge was essential for guiding, evaluating, and correcting AI suggestions. I’ll also share examples of learning through collaboration, including moments when I operated outside my expertise (such as diagnosing and recovering from a security breach) and how existing competencies were crucial to achieving desired results. The session will identify transferable principles that apply across disciplines beyond coding: how expertise shapes better prompts, why productive struggle matters, and what foundational competencies enable useful AI collaboration.
Participants will discover implications for teaching and learning with AI: why students need disciplinary competencies before they can collaborate effectively with AI, and how we can teach AI literacy as an extension of expertise rather than a replacement for learning. Through actual AI conversation examples from both coding and non-coding contexts, the session will reveal collaboration patterns across different types of work, emphasizing that AI tools amplify expertise rather than replace it, with practical takeaways for cultivating this approach in teaching and learning.
