Relational Autotheory White Paper: AI Literacy Through Sustained LLM Dialogue: A Practical Method for Supporting Human Thinking Without Outsourcing Responsibility

Relational Autotheory (RA) proposes a conceptual framework and practical method for supporting human thinking through sustained, ethically governed dialogue with large language models. RA treats long-horizon interaction as a potential cognitive scaffold that can stabilize context and conversational structure over time, enabling higher-order reasoning, creative synthesis, and meta-observation without outsourcing responsibility. This white paper distinguishes training-layer alignment from interaction-layer governance and offers practices designed to preserve human agency and epistemic integrity, including scope boundaries, correction norms, resistance to unearned affirmation, and active monitoring for drift and hallucination. It is presented as a working contribution intended for critique, refinement, and future empirical study. This white paper includes: Core definitions and conceptual model Governance practices and dialogic norms Concrete examples for professionals and avid LLM users Annotated references and internal navigation for study and teaching