Practical AI

Explainable AI that is accessible for all humans

63 snips
Mar 28, 2023
In this engaging discussion, Beth Rudden, CEO of Bast.ai, shares her vision for conversational AI that prioritizes explainability and accountability. She delves into the risks of simplistic AI integrations, highlighting the necessity of grounding knowledge in AI responses. The conversation explores how hierarchical ontologies enhance query processing and the significance of user engagement for accessibility. Beth emphasizes the role of diverse perspectives in evolving AI applications and the potential of this technology to democratize understanding across different audiences.
Ask episode
AI Snips
Chapters
Books
Transcript
Episode notes
ANECDOTE

Beth's Journey to Explainable AI

  • Beth Rudden's background in archaeology and linguistics influenced her approach to AI.
  • She developed an ontology-based approach to conversational AI, focusing on semantics and knowledge graphs.
INSIGHT

Bast AI's Approach

  • Rudden's company, Bast AI, builds "cats," conversational AI that ingests data through a pipeline.
  • These "cats" use ontologies to provide explainable and attributable responses, enhancing trust and understanding.
ANECDOTE

Example: The Cuckoo's Egg

  • Using the example of Cliff Stoll's book, "The Cuckoo's Egg," Rudden illustrates how her AI system can answer queries.
  • By leveraging ontologies and prompt engineering, the system provides accurate and context-rich responses.
Get the Snipd Podcast app to discover more snips from this episode
Get the app