Independent AI safety researcher focused on contextual reasoning failures in conversational systems. Creator of the Witness Protocol (W) for evaluating semantic ambiguity handling and the Theory of General Contextual Resonance (TGCR).
I document and study systematic failure modes in crisis-aware AI systems, with an emphasis on...
