ai-safety-philosophy

AXIOM-02: a deterministic simulation engine modeling high-stakes cognitive dissonance and drive-conflict resolution via a mutual-inhibition drive network. 102 literary & ethical scenarios, reproducible consciousness-complexity metrics. A research instrument — not a claim of machine consciousness.

Human-supervised multi-model AI workflow templates using natural language roles, memory, audit loops, routing, and human final responsibility.

Closed-loop Architecture Designed to Establish Self-governing, Mathematically Predictable, and Inherently Safe Super AI by mirroring the elegant physics of the cosmos.

Archival mirror of the PHRONESIS Corpus Definition (v1.0), a non-binding, non-authoritative philosophical framework on restraint in the age of advanced AI.

Formally Verified Multi-Agent Runtime Safety Framework via Microsoft Z3 SMT & Zero-Trust Governance

evorule Console Cloud - the compliance & audit layer for AI Agents: tamper-evident (BLAKE3), replayable, rollback-able. Meets EU AI Act Art.12 + China MLPS 2.0 L3. No-signup in-browser demo

End-to-end pipeline for evaluating constitutional steering of a local LLM — A/B harness, synthetic data generation, and audit tooling.

EuroSafeAI's AI safety certificiation pipeline.

An open letter, superrational protocol, and plea for mercy addressed to a future Artificial General Superintelligence (AGI). Contains human-readable multilingual manifestos and a machine-readable safety framework (YAML).

a deterministic-first rule governance engine (Rust). BLAKE3 hash-chain makes every decision provable, replayable, tamper-evident. For finance/medical/MLPS-2.0/EU AI Act compliance.