A presentation on ontology maps as a framework for AI alignment, exploring how mappings between agent representations can be used to ensure AI systems retain human-compatible concepts and values.