AOS Constitutional AI Thesis & Foundational Documentation

Author: Eugene Christopher Salvatore | Subject: Deterministic AI Governance vs Probabilistic Alignment

Thesis Abstract

This thesis establishes that probabilistic AI alignment techniques (RLHF, system prompt tuning, constitutional training) are insufficient for autonomous agent governance in production environments with persistent side effects.

We propose and formally prove the deterministic constitutional paradigm: decoupling cognitive generation from physical execution through an external, verifiable, cryptographic gate.