AI Safety Proposal: Systemic Interdependence Evaluation (SIE) Can an AI system be highly intelligent while failing to understand the consequences of its actions on interconnected systems? As AI becomes increasingly autonomous, I believe safety evaluations…