Explainable AI that is accessible for all humans

Semantic Design

Beth Rudden, founder of Bast.ai, emphasizes the importance of semantic and ontological designs in conversational AI. She explains how their technology, named after the Egyptian cat goddess, integrates a data pipeline to ingest user data and map it into a knowledge graph, enhancing natural language understanding (NLU) 1. This approach allows AI to generate meaningful interactions by understanding the user's reality through language 2.

An ontology is the study of the nature of your reality based on the language that you use.

--- Beth Rudden

By leveraging ontologies, Beth's team can create conversational AI with lineage and provenance, ensuring that responses are grounded in context and relevant to the user's queries.

   

Entity Management

Entity management in conversational AI is crucial for tailoring responses and maintaining relevance. Beth describes how entities are processed through a data pipeline, allowing the AI to respond accurately by referencing the ontology 3. This method ensures that AI interactions are explainable and transparent, as users can see the source of the AI's responses.

We wanted to show people how they could ingest a paragraph and then be able to communicate with conversational AI.

--- Beth Rudden

Beth also highlights the importance of testing AI systems to ensure reliability and trustworthiness, using innovative methods like having AI "cats" test each other 4.

   

User Engagement

User engagement in conversational AI involves ensuring interactions are both interactive and meaningful. Chris Benson points out the challenge users face in not attributing intelligence or consciousness to AI systems, which are essentially "stochastic parrots" that mimic language 5. Beth advises users to engage with AI by asking questions about familiar topics, helping them understand AI's capabilities and limitations.

It generates transformations. It does not have the understanding to have consciousness or sentience.

--- Beth Rudden

Daniel Whitenack shares insights on how users, even those unfamiliar with AI, often engage responsibly by using AI to enhance their tasks rather than relying on it completely 6.