Building Trustable AI: Navigating Governance, Documentation, and Human Judgment with Koo Ping Shung

In this episode of the API Connections Podcast, Jon Scheele continues his deep dive into the complexities of AI governance with data scientist Koo Ping Shung. Following their discussion on AI bias and transparency, they explore the challenges organizations face in implementing robust data practices and the pivotal role of documentation in AI governance. Koo emphasizes the need for building trustable AI, arguing that trust, rather than strict adherence to governance or ethics alone, is the key to successful AI deployment. Together, they delve into the nuances of documentation, knowledge management, and the human element in AI, offering insights into creating AI systems that are both reliable and aligned with human values.Sections:Challenges in Using Data and AIThe Importance of Documentation for AI GovernanceTesting Models and Creating ModelsThe Challenge of DocumentationUsing AI for Documentation and Knowledge ManagementContinuous Improvement and Data QualityBuilding Trustable AICommunication and Decision-Making in AI GovernanceThe Limitations of AI Ethics