RICCI
RICCI is a PostgreSQL detection architecture that identifies coordination risk in communication systems from interaction topology alone. Rather than scanning message content or matching keywords, it operates on the geometric structure of how accounts interact: timing, relational shape, embedding curvature, and coupling dynamics across the interaction graph.
The system is structurally incapable of reading raw messages. Its entire detection surface lives in geometry, and the output is designed to be both auditable and classificatory.
Twelve independent detectors, derived from first principles, span four categorical failure modes in coupled systems: rigidity, fragmentation, inflation, and distortion. Each detector returns a normalized severity score, geometric evidence arrays, and human-readable mathematical justification for the thresholds it crossed.
github.com/diesel-black/ricciThe Cubic Seam
The coupling operator governing coordinated behavior in a communication graph has a polynomial factorization whose order determines the detection architecture required. At the cubic threshold, standard spectral tools fully decompose the field. Above cubic, the signal survives only in the coupling structure between accounts.
A single diagnostic measures where any interaction dataset sits on this boundary. Compute the singular value decomposition of the coupling matrix and take the ratio of the leading singular value squared to the total Frobenius norm.
This ratio is computable from any existing interaction dataset in a few lines of Python, with no threshold calibration required.
Per-account scoring captures available signal
Signal exists only in the graph's coupling structure
Trajectory
I learned to read fields before I learned to read people. HAM radio as a teenager, then electromagnetic systems in the Marine Corps, where standing between two synchronized diesel generators and feeling them go silent taught me what phase-locked coherence feels like in your chest.
Seventeen years of felony investigations as a police officer and detective built the rest: tens of thousands of investigative interviews, extracting signal from adversarial noise, watching how coordinated deception bends a story's geometry the same way every time, regardless of content.
Inside The Black Box exists to formalize the pattern recognition that made the investigative work effective, building mathematical infrastructure to make the dynamics of complex systems transparent and tractable.
Paper
Semantic Relativity: A Geometric Field Theory for Observer-Dependent Meaning
The paper begins from seven axioms about how meaning behaves in any system capable of maintaining it and derives the coupled field equations, Lagrangian structure, twelve orthogonal detection signatures, and four structural falsifiers that govern the theory.