Agentic - Ethical AI Leadership and Human Wisdom

#12 - The Only Realistic Path to Safe AI: Exidion’s Living-University Architecture

Listen Later
Exploring AI alignment, bias mitigation, and human-centered AI safety architectures
  • Epistemic & psychometric layer: Rules and measurements that check whether an AI’s reasoning stays oriented, coherent, and aligned with human values.
  • MoE (Mixture of Experts): Many small specialist models coordinated by a router, instead of one all-purpose model.
    • -RAG (Retrieval-Augmented Generation): The model looks up verified sources at answer time, instead of “guessing” from memory.
  • Distillation: Compressing the useful behavior of a large model into a smaller, efficient model.
  • Agency drift: When a system’s behavior starts to pursue unintended strategies or goals.
  • Governance-legible: Decisions and safety controls are traceable and explainable to auditors, boards, and regulators.
    • ...more
    View all episodesView all episodes
    Download on the App Store
    Agentic - Ethical AI Leadership and Human WisdomBy Christina Hoffmann - Expert in Ethical AI and Leadership