The bottleneck has moved from the model to context: connecting intelligent systems to the right data, with the right meaning, wherever it resides. Over dinner, we'll discuss with peers the barriers that stand between ambition and value: accuracy, cost, security and governance.

 

1) Meaning first: semantics and ontologies

Giving models the right context, and only the context they need.

The Discussion: To what extent do semantics and ontologies provide the foundation for accurate, trustworthy results, and how can they reduce token cost?

2) Securing sensitive data as models and agents gain access to it

The Discussion: How do you protect sensitive data as models and agents gain access to it?

3) Governance when agents act on enterprise data

The Discussion: What does governance look like when agents act on enterprise data, and what role does guided AI play in keeping them on track?