Interaction Patterns: Elements of Effective AI Collaboration By Dan Rapp and Claude
A practical guide to interaction patterns for working effectively with AI applications — Claude, Claude Code, agentic workflows, and the emerging class of AI-powered tools.
Inspired by the Gang of Four’s Design Patterns: Elements of Reusable Object-Oriented Software, this book gives practitioners a shared vocabulary for recurring collaboration problems in human-AI work.
Most people interact with AI tools either too passively (accepting whatever the model generates) or too prescriptively (trying to specify everything upfront). There’s a middle ground — interaction patterns — that dramatically improves output quality.
Just as a novice and expert programmer both have access to the same language features but the expert recognizes recurring problem shapes and applies proven approaches, a skilled AI practitioner recognizes when to interview vs. generate, when to constrain vs. explore, when to validate vs. trust.
The book’s central thesis is Complementary Asymmetry — effective human-AI collaboration is a dynamic balance of complementary strengths. The human provides judgment, recognition, and direction. The AI provides volume, structure, and articulation. Each contains a seed of the other.
14 chapters cover 20 patterns organized in four parts: The Thesis, Discovery, Environment, and Scaling.
Read the full text (password protected)