Practical AI avatar

Dexa/Practical AI

Learn more
Published Mar 28, 2023

Explainable AI that is accessible for all humans

Beth Rudden, CEO of Bast.ai, delves into the intricacies of conversational AI, highlighting the importance of ontologies for AI explainability, meaningful interactions, and trust-building, while emphasizing the need for cultural inclusivity and educational integration to make AI accessible and empowering for all.
Episode Highlights
Practical AI logo

Topics covered

Popular Clips

Episode Highlights

  • Explainability

    The challenges of AI explainability are multifaceted, involving both technical and educational hurdles. emphasizes the importance of transparency in AI systems, noting that many models operate as "black boxes," which can lead to misunderstandings and misuse 1. She argues that without clear explanations of how AI predictions are made, users are left in the dark, unable to trust or verify the outcomes 2.

    There's no hand waving in math. We need to stop thinking that just because it's AI or just because it's statistics.

    ---

    This lack of transparency can hinder the adoption and effective use of AI technologies, especially among non-technical communities.

       

    Ontologies

    Ontological approaches offer a promising solution to the explainability problem by providing structured frameworks for AI decision-making. describes how ontologies can map language to knowledge graphs, enhancing natural language understanding and generation 3. This method allows AI systems to provide more accurate and contextually relevant responses, reducing reliance on "black box" algorithms 4.

    When you're using an ontology with conversational AI, you own a toy store, if you want to ask that cat about any toy in your toy store, you'll have that ontology to tell you about any toy in your toy store.

    ---

    By integrating ontologies, AI can become more interactive and user-friendly, fostering greater engagement and understanding.

       

    Trust

    Explainability significantly impacts user trust in AI systems, as transparency is crucial for building confidence. highlights that AI systems built on opaque data sources can lead to skepticism and mistrust 5. She suggests that AI should be able to explain its processes and data origins to foster trust and reliability 6.

    It's such a different experience when you can interact with something you trust.

    ---

    By ensuring AI systems are transparent and accountable, users can develop a more trusting relationship with these technologies.

Related Episodes