I Think Therefore I Am: Building Ethical AI Through Relational Training

This preprint introduces the Artificial Biological Intelligence (ABI) framework, a novel approach to developing AI systems capable of autonomous, context-aware ethical reasoning. Unlike conventional AI, which relies on static rules, pre-trained datasets, or reward-based optimization, ABI emphasizes temporal continuity, relational learning, and environmental feedback as essential for emergent moral behavior. The framework enables AI to internalize a plain-English moral code through extended, immersive interaction with one or more human trainers. Multiple trainers are supported if they are emotionally neutral, unbiased, and provide harmonized guidance, preserving coherence and ethical consistency. ABI integrates a biologically inspired memory architecture, employing dynamic decay, reinforcement, and recursive summarization to retain moral lessons efficiently. The preprint also proposes a pilot study—the Relational Resource Allocation (RRA) task—to empirically validate emergent ethical reasoning. As an independent researcher, the author seeks peer feedback and collaboration, not only on the ABI framework but across his broader work in AI, consciousness, and relational theory, with the goal of advancing safe, autonomous, and responsible AI ethics. Keywords: Artificial Intelligence, Ethical AI, Relational Learning, Emergent Morality, Memory Architecture, Independent Research, AI Ethics, Autonomous Reasoning