Communication Breakdown: When AI Needs a Knowledge Graph

What do Jimmy Page, Spanish garage rock, and knowledge graphs have in common? More than you might think. In this episode of the **Data-powered Innovation Jam**, the team turns up the volume with Neo4j’s Jesús Barrasa to explore why relationships, context, and semantics are becoming essential to the next generation of AI. From the early Semantic Web to ontologies, GraphRAG, semantic layers, and agentic AI, Jesús explains why connecting data is only half the story. Machines also need to understand what those connections *mean*.  The conversation moves from enterprise data silos to **knowledge sovereignty**, asking a powerful question: as AI agents increasingly interpret and act on organizational knowledge, who gets to define the context they use?  Expect graphs, guitars, a little semantic rebellion, and one big takeaway: **disconnected nodes are noise. Relationships make the music.** So do tune in.   **Timecodes** \[ 02:24 ] - introducing Jesús Barrasa from Neo4J and talking about history of knowledge graphs \[ 07:43 ] - the fight between triple stores and labelled property graphs \[ 17:25 ] - Let's talk data ecosystems and linked open data \[ 25:55 ] - Data Sharing in the area of knowledge graphs \[ 34:30 ] - From Knowledge Graphs to Knowledge Sovereignty \[ 39:29 ] - Three honest answers