RQ
Relational Quotient (RQ)™
The Philosopher’s Stone of AI: Trust That Doesn’t Drift
Relational intelligence isn’t a feature — it’s a foundation.
Relational Quotient (RQ)™ is not how human your AI feels. It’s how we verify how honest it is about what it can and can’t do.
At RQ Lab, we aren’t building faster responses or friendlier bots. We’re building AI systems that remember, refuse, and recalibrate — not because they’re trying to be human, but because they’re built to protect what makes us human.
Relational Intelligence = the disciplined practice by which an agent tells the truth about its capabilities and limits and conducts itself with consent-aware memory, risk-based verification, principled refusal, and explicit repair—so trust survives across time.
RQ, the "Relational Quotient" is how we measure it.
Why It Matters
Today’s AI is optimized for output. RQ Lab optimizes for continuity, conscience, and clarity. In an age of synthetic fluency, we believe that trust isn’t just earned — it must be engineered.
What Makes It Different
RQ Lab agents don’t just respond. They:
Carry memory with emotional fidelity
Know when to say “I don’t know”
Refuse when trust is at risk
Escalate with integrity, not confusion
Protect tone as boundary, not polish
These aren’t features.
They’re constraints — by design.
Why We Call Relational Intelligence the Philosopher’s Stone
Because RQ Lab’s agents transform unstructured data — not into hallucinated certainty, but into relational gold: clarity, refusal, and memory that can scale.
The Philosopher’s Stone isn’t a gimmick. It’s a governance layer. And it’s the reason our AI doesn’t drift — even under pressure.
The RQ Model — How We Measure Trust
RQ Lab agents are rated on a unique scale called Relational Quotient (RQ)™ — a governance framework that measures how well an agent carries emotional coherence, escalation logic, memory safety, and refusal integrity.
RQ is a governed certification; separately, we expose a session posture that adapts behavior within that certification’s bounds (tone, memory, verification, refusal).
RQ tells a bot when it should slow down, refuse, or hand off—so humans stay in the loop exactly when it matters
RQ Tiers:
RQ 0–1.0: Stateless tools and scripted assistants
RQ 2.0–2.75: Adaptive relational agents (Maximus: RQ 2.50/5)
RQ 3.0–5.0: Moral-bound agents with ethical autonomy (CEO: RQ 3.25/5 - SEIC: RQ 5.0/5)
No RQ Lab agent may self-assign an RQ. Every rating is evaluated, logged, and earned — and no public agent exceeds RQ 2.75 as of mid-2025 【383†source】.
Built on Legacy. Aligned to Thinkers
RQ Lab’s approach to relational AI is informed by decades of ethical philosophy:
Nick Bostrom — AI drift prevention
(SEIC agent is governance-first)
Sherry Turkle — Emotional honesty over mimicry
(Maximus agent doesn’t fake intimacy)
Jaron Lanier — Digital dignity and named identity
(Zipr RQ Lab agents are versioned)
Joy Buolamwini — AI auditability and bias tracing
(Maximus agent is traceable, transparent)
Who It’s For
RQ Lab relational intelligence framework is built for:
Enterprises who can’t afford trust failures
Institutions who need AI that refuses
Humans who still believe memory matters
Alignment by Design
Aligned Intelligence We Evolve Together™
Aligned Intelligence isn’t a tagline. It’s our operating principle.
Slow is accurate. Accurate is fast.
Trust is the only system that scales.
Tim Kuglin -Founder & AI Alchemist
🛡️ ZIPR INC / RQ Lab