Finbarr Murphy discusses RAGgy Meshes. Or is it Meshy RAGs...
Yesterday, Keir Starmer set out his vision for AI in government—promising a digital revolution that will make services more efficient and save billions. But as I said before, AI isn't a shortcut to cost-cutting—it's a tool that, when done right, can transform public services. So, let's imagine for a second that all the barriers are magically removed. Data governance? Perfect. Data access? Seamless. Security, quality, and interoperability? All sorted. If that were the case, how should government actually 'do AI'? It begins with the letter "t." Below, I've written about why open, explainable, and accountable AI must be the foundation of any public sector AI strategy. LLMs, RAG, and Data Mesh offer incredible potential—but only if we build AI that policymakers, civil servants, and citizens can trust. If we want AI to work for the public good, we need open models, transparent data pipelines, and an AI strategy focused on explainability at every level. This is how AI in government should be done. Read more below. Modular Data #AI #PublicSector #DataGovernance #ExplainableAI #Transparency #DigitalTransformation #RAG #LLMs #DataMesh