COrigami: AI Pipeline Co-Designs Flat-Foldable Visual Origami.
Summary
This paper introduces COrigami, an AI-driven pipeline that assists in designing flat-foldable, visually recognizable origami from natural language descriptions. The system integrates semantic stick figure generation, base packing, crease pattern solving, shaping, and reinforcement learning with aesthetic evaluation, acting as a collaborative tool for artists.
Why it matters
This technology showcases how AI can bridge the gap between strict mathematical constraints and subjective artistic expression, opening new avenues for design and co-creation in fields requiring both precision and aesthetics, from product design to architecture.
How to implement this in your domain
- 1Explore integrating AI co-design pipelines into creative workflows for products requiring both functional constraints and aesthetic appeal.
- 2Develop natural language interfaces for design tools to allow users to describe desired outcomes intuitively.
- 3Implement autonomous evaluation loops using AI to critique designs based on predefined aesthetic or functional criteria.
- 4Utilize reinforcement learning to iteratively refine designs, optimizing for multiple objectives simultaneously.
- 5Collaborate with artists and designers to understand how AI tools can best augment their creative processes, providing structural foundations for human refinement.
Who benefits
Key takeaways
- AI can successfully co-design physical art that adheres to strict geometric rules and subjective aesthetics.
- The COrigami pipeline integrates multiple AI techniques for an end-to-end design process.
- Autonomous aesthetic evaluation and reinforcement learning are key to refining artistic AI outputs.
- AI can serve as a powerful collaborative assistant for human artists and designers.
Original post by Tom Zahavy, Shaobo Hou, Thomas Tumiel, James Doran, Francesco Faccio, Xidong Feng, Alex Havrilla, Igor Khytryi, Chenglei Li, Lisa Schut, Vivek Veeriah, Arijan Abrashi, Micha{\l} Kosmulski, Robert J. Lang, Nick Robinson, Brandon Wong, Marcus Chiam, Gloria Fang, Satinder Singh
"arXiv:2606.26299v1 Announce Type: new Abstract: While generative AI has achieved remarkable success in solving problems with verifiable solutions, generating physical art that satisfies both strict geometric constraints and subjective visual aesthetics remains a challenge. This p…"
View on XOriginally posted by Tom Zahavy, Shaobo Hou, Thomas Tumiel, James Doran, Francesco Faccio, Xidong Feng, Alex Havrilla, Igor Khytryi, Chenglei Li, Lisa Schut, Vivek Veeriah, Arijan Abrashi, Micha{\l} Kosmulski, Robert J. Lang, Nick Robinson, Brandon Wong, Marcus Chiam, Gloria Fang, Satinder Singh on X · view source
Want to go deeper?
Turn these trends into skills with Learnijoy's hands-on AI & tech courses.
Explore coursesMore in AI Engineering & DevTools
AI-Powered Development Workflow Integrates Multiple Models
A new development workflow leverages various AI models like Grok 4.3, GPT-5.5, and Opus 4.8 for distinct stages including research, planning, coding, testing, and debugging. This structured approach aims to optimize the software development lifecycle.

Proposing AI Usage Transparency for Credible Commentary
The author suggests a requirement for individuals and organizations to publish their percentage of frontier AI usage at work and personal usage. This transparency would establish credibility before commenting on AI's utility.
MCP and A2A Protocols Standardize Agentic Internet Development
The Model Context Protocol (MCP) and Agent-to-Agent (A2A) Protocol are standardizing how AI agents discover tools, call services, and coordinate across systems. Understanding these protocols is crucial for developers building agent-compatible infrastructure.