Research Synthesis
AI tools help synthesize information from classical texts, academic references, and scholarly articles. Published records expose their sources so claims can be checked against primary or authoritative secondary material.
ZHAIBIAN Knowledge Archive
AI Transparency
How artificial intelligence is used — and not used — in the ZHAIBIAN archive.
Principle
AI-assisted and source-visible
ZHAIBIAN uses artificial intelligence in research synthesis, drafting, organization, and archive maintenance. The archive publishes its sources and explains its validation process so readers can assess the work and continue the inquiry.
AI systems assist with research synthesis, drafting, content structuring, and knowledge graph maintenance. Before publication, records must pass automated checks for required sources, metadata, structure, duplication, and knowledge-graph relations. These checks do not constitute human expert review.
Where AI is used
Assisted processes
Content Structuring
AI assists in organizing drafts according to archive templates for Wisdom, Answers, Thinkers, Books, Philosophies, Quotes, Topics, and Collections.
Knowledge Graph
AI helps identify and map relations between thinkers, books, philosophies, and wisdom concepts. Editors validate relation quality and remove low-value connections.
Quality Monitoring
Automated scripts check frontmatter, sources, query ownership, semantic duplication, relation integrity, metadata, and build health before publication.
Where AI is NOT used
Editorial boundaries
- ×Unsupported publication. Records that fail source, structure, duplication, metadata, or relation checks are not eligible for publication.
- ×Source verification. AI does not verify the accuracy of quotations, citations, or historical claims. Source references are exposed on each record for independent verification.
- ×Opinion or interpretation. AI does not express editorial opinions. All interpretive content reflects the editorial team's judgment.
- ×Strategic decisions. AI does not decide what topics to cover, what clusters to build, or what content to prioritize.
Commitment
Source-first, always
Our commitment is to the reader. Every record lists its sources. Every quotation is traceable. Every connection in the knowledge graph is intentional. If you believe an AI-assisted process has introduced an error, we encourage you to report it through our corrections process.