Introduction to When Large Models Meet Liu Yao: An Exploration of AI's Logical Analysis in Wen Wang Gua Divination

Original Author: goldmanau Source: GitHub (Link) Publication Time: June 2, 2026

This study explores the application potential of generative AI in the field of Liu Yao (Wen Wang Gua) divination, testing large language models' ability to understand rules, perform logical reasoning, and make fuzzy judgments in this highly structured traditional metaphysical system. It focuses on core questions such as whether AI can master Liu Yao rules, handle empirical content, and achieve interpretable analysis, while also discussing the value and boundaries of combining technology with traditional culture.

Research Background: Structured Characteristics of Liu Yao Divination

Liu Yao divination is based on the Zhou Yi (Book of Changes). It constructs hexagrams through divination methods (e.g., coin tossing) and deduces outcomes by combining rules like the five elements' generation/restriction and six kinships. It has structured features including a fixed symbol system (Eight Trigrams, Sixty-Four Hexagrams, etc.), clear operation rules (five elements' generation/restriction, prosperity/decline judgment), and standardized processes (hexagram installation, Yongshen checking, etc.), making it an ideal scenario to test AI's logical and fuzzy reasoning abilities.

Research Motivation: Core Questions

1. Can large models accurately understand and apply the Liu Yao rule system?
2. How does AI perform in fuzzy reasoning tasks?
3. Can traditional metaphysical knowledge be formally taught to AI?
4. Is AI analysis interpretable and consistent?

These questions touch on the deep relationship between AI, traditional culture, and human experience.

Core Mechanisms of Liu Yao Divination

1. Hexagram Generation and Installation: Determine hexagrams through specific methods (e.g., coin tossing);
2. Shi-Ying Positioning: Shi Yao represents the subject, Ying Yao represents the object—key to judging the situation;
3. Six Kinships Configuration: Assign six kinships (parents, official ghosts, etc.) to Yao positions based on the hexagram palace's five elements and Earthly Branches;
4. Yongshen Selection: Choose the corresponding Yongshen based on the nature of the question (e.g., select Wife-Cai Yao for financial matters);
5. Prosperity/Decline and Moving Changes: Analyze the Yongshen's prosperity/decline under the influence of the moon and sun, and changes in moving Yao.

AI Test Design

• Knowledge Input: Structured input of Liu Yao basic theory, hexagram installation algorithms, analysis rules, etc.;
• Test Cases: Cover multiple types of questions (career, relationships, health, etc.);
• Evaluation Dimensions: Rule compliance, logical consistency, knowledge accuracy, conclusion rationality.

AI's Performance Highlights

1. Rule Understanding Ability: Can correctly identify hexagrams, calculate five elements' generation/restriction, install hexagrams and configure six kinships per process, and select Yongshen—indicating structured traditional knowledge can be effectively learned by AI;
2. Reasoning Transparency: Provides clear steps (e.g., Shi-Ying determination, Yongshen selection reasons), with better interpretability than traditional divination.

Limitations

1. Lack of Experience: Struggles to grasp subtle empirical differences like "prosperity" and "decline";
2. Fuzzy Reasoning Difficulty: Unstable performance when balancing multiple factors;
3. Cultural Context Bias: May have deviations in grasping cultural connotations;
4. Insufficient Creativity: Hard to flexibly apply like human diviners.

Technical Significance

1. Exploration of AI's Capability Boundaries: Tests AI's performance in "semi-structured" tasks (rules + experience), consistent with observations in legal and medical fields;
2. Cultural Digitalization: Formalizing Liu Yao rules helps preserve and spread traditional knowledge;
3. Interdisciplinary Demonstration: Provides a reference for combining AI with traditional humanities.

Application Prospects

1. Learning Assistance: Helps learners verify hexagram installation, understand five elements relationships, and practice Yongshen selection;
2. Research Tool: Batch analysis of hexagram cases, verification of theoretical consistency;
3. Decision Reference: Provides a structured analysis framework to assist decision-making.

Technical Limitations

• AI analysis is based on rule simulation, with no supernatural abilities;
• Prediction accuracy has no scientific verification, for academic reference only;

Cultural Sensitivity

• Respect traditional cultural values and beliefs, avoid over-sanctifying technology;
• Clarify the boundaries between academic research and cultural practice;

Ethical Boundaries

• Do not encourage relying on AI for major decisions;
• Do not claim AI has the ability to predict the future;
• Maintain research objectivity and criticality.

Future Research Directions

1. Model Enhancement: Fine-tune open-source models, introduce multimodality, integrate knowledge graphs;
2. Methodology Improvement: Establish standardized datasets, develop human-machine collaboration models, quantify uncertainty;
3. Cross-domain Expansion: Apply to other traditional prediction systems (BaZi, Zi Wei Dou Shu, etc.) to explore cultural commonalities.

Conclusion

This study examines the combination of AI and traditional knowledge with an open attitude. It neither proves AI can replace humans nor verifies the effectiveness of metaphysics, but seeks possibilities for dialogue between technology and culture. The interdisciplinary attempt itself is valuable, reminding us to look back at traditional cultural accumulation while advancing technology.