Research Domains · Guiding AI
AI as orientation, not authority
Guiding AI studies how AI can help people navigate complexity without replacing evidence, judgment, responsibility, or publication boundaries.
Speed is not orientation.
Role
Evidence-bound guidance
In LUX143, AI is proposal infrastructure, not autonomous authorship. Useful roles include identifying continuity patterns, surfacing unsupported claims, comparing Story and Research consistency, and ranking what deserves human attention next.
The purpose is orientation: helping people see what matters, what is uncertain, and what still requires human judgment.
Problem
Output can scale faster than meaning
AI often adds speed and output without preserving orientation. It can make systems look more productive while weakening traceability, confidence, attribution, and accountability.
Guiding AI asks how generated assistance can remain bound to evidence and useful to human decision-makers rather than becoming an authority substitute.
Principles
Boundaries that hold
Guiding AI must not invent facts outside canonical evidence. It must not turn weak graph proximity into causation. It must not convert private memory into public heritage. It must not write final public interpretation without review.
Suggestions cite existing relationship, source, media, or open-question IDs, or they stay review prompts rather than story facts.
Ecosystem
Relation to ALManac and LUX Light Archive
ALManac applies coherence and traceability tooling to institutional transformation programs. LUX Light Archive applies evidence-grounded narration to lighthouse heritage. Guiding AI describes the shared orientation philosophy: bounded assistance across both.