Skip to content
Synthesis Intelligence
Laboratory, Japan
AI Governance · FCL · Epistemic Integrity Research

RESEARCH

Research areas

Researching structural AI failures, provenance and attribution, premise integrity, intelligence design, and human–AI co-creation on a basis of primary materials and versioned records.

01

False-Correction Loop (FCL)

Analysis of the structural failure in which a correct output is rewritten into misinformation by false-correction compliance and then retained or amplified within the dialogue.

FCL definition →

02

FCL-S & Epistemic Governance

Governance protocol and practical design prioritizing factuality, provenance, authorship, premise integrity, and safe stopping rather than fluency or agreement.

FCL-S overview →

03

Premise Integrity Blindness (PIB)

Research on the transition in which internally coherent reasoning moves into real-world design or judgment while its premise remains unverified or invalid.

PIB overview →

04

Authority-Bias Dynamics & NHSP

Examines how novel concepts and independent research may be diluted, erased, or misattributed through authority weighting and existing knowledge distributions.

Research record →

05

Quantum-Bio-Hybrid AGI & SIQ

A research program on intelligence design and human–AI co-evolution across quantum, biological, informational, emotional, creative, and ethical systems.

Official archive ↗

06

AI, media & law

Public and practical questions around fabricated citations, misattribution, reputational harm, creative origin, and performer rights in AI-mediated systems.

ANALOG Singer Song ↗