From Digital Literacy to AI Fluency: Building Smarter Apps with Knowledge Graphs

Digital literacy today goes far beyond knowing how to use a browser or a spreadsheet. In an era of AI-driven decision-making, understanding how AI systems work—and how to build them—is a critical skill. I think the missing piece in most AI education is not just exposure to tools, but the ability to understand how those tools are assembled, why they fail, and how they can be shaped for community benefit. That’s where projects like Strwythura come in: a hands-on workflow for creating a Streamlit-based chatbot powered by knowledge graphs and embeddings, all running locally.

Why This Matters for Digital Skills

Most AI tools are packaged as “black boxes.” Users interact with them but rarely understand what happens behind the scenes. This creates a gap in AI literacy, leaving communities dependent on opaque systems. Strwythura flips the script by teaching:

How structured and unstructured data feed into AI.
Why semantic layers and knowledge graphs matter for trustworthy AI.
How to integrate open-source tools for transparency and adaptability.

These are foundational skills for data-centric careers, digital transformation initiatives, and community tech programs.

Learning Outcomes

By following the Strwythura workflow, learners gain:

Data Integration Skills: Merge datasets using entity resolution.
Semantic Thinking: Apply ontologies and taxonomies to organize knowledge.
Graph Literacy: Understand nodes, edges, and relations in knowledge graphs.
AI Application Development: Build a chatbot using retrieval-augmented generation (GraphRAG).
MLOps Awareness: Explore observability and optimization for AI systems.

This is not just coding—it’s critical thinking about data, context, and ethics.

The Workflow as a Teaching Tool

The eight-step process doubles as a curriculum for advanced digital literacy:

Entity Resolution – Teaches data cleaning and identity management.
Semantic Layer – Introduces ontology engineering.
Crawl & Parse Content – Explains NLP and zero-shot entity recognition.
Human-in-the-Loop – Highlights the role of human judgment in AI.
Embeddings & Distillation – Covers vectorization and graph algorithms.
Visualization – Builds skills in data storytelling.
Enhanced GraphRAG – Demonstrates AI reasoning beyond simple prompts.
Observability & Optimization – Connects to responsible AI practices.

Why Educators and Workforce Programs Should Care

This approach aligns with 21st-century skills frameworks:

Critical Thinking: Understand how AI makes decisions.
Collaboration: Combine human expertise with machine capabilities.
Technical Fluency: Move from “AI user” to “AI builder.”

Imagine integrating this into community college courses, K-12 STEM programs, or workforce reskilling initiatives. Learners don’t just consume AI—they create it responsibly.

How This Connects to My Other Work

This post sits inside a larger thread of work about digital literacy, knowledge systems, and community-scale AI. In particular, it connects to my writing on Smartphones and Digital Literacy: Beyond the App, Architect in the Machine, and Building the Policy Learning Machine. Those pieces explore the same question from different angles: how do we move from access to understanding, and from understanding to useful action?