The LinkMap
98 terms and the lines between them.
Understand the system
behind the headline.
Fresh from the desk.
What is MCP (Model Context Protocol)? A beginner's guide to the "USB-C for AI"
AI assistants are most useful when they can reach your real files, tools, and data — but wiring each one up by hand is slow and brittle. MCP is a shared standard that lets any AI app plug into any tool through one common connector, the way USB-C lets one cable charge almost anything. Here is what that means in plain language, why it caught on so fast, and where it still trips people up.
What is fine-tuning? A beginner's guide to teaching an AI your way
When a general AI model needs to sound like your brand, follow your format every time, or handle a task it keeps fumbling, fine-tuning is one way to teach it. It takes a model that already knows a lot and trains it a little more on examples of exactly what you want. Here is what that means in plain language, how it differs from prompting and RAG, and when it is — and is not — worth the trouble.
What are AI embeddings? A beginner's guide to how AI understands meaning
When an AI finds the right document, recommends a similar product, or answers from your company files, an embedding is usually doing the quiet work underneath. It turns words, images, and whole sentences into lists of numbers so a computer can measure how close two meanings are. Here is what that means in plain language, why "search by meaning" suddenly works, and where it still trips up.
What is tool calling (function calling)? A beginner's guide to how AI assistants get real work done
A chatbot can write you a paragraph, but on its own it cannot check today''s weather, look up your order, or do the maths reliably. Tool calling is how it reaches outside the conversation to use real tools — and it is the quiet machinery behind almost every AI "agent." Here is what that means in everyday language, how the back-and-forth works, and where it still trips up.
What is RAG (retrieval-augmented generation)? A beginner's guide to AI that answers from your own documents
A plain chatbot only knows what it learned during training, so it cannot answer questions about your company handbook or last week''s notes. RAG fixes that by fetching the right snippets first and handing them to the AI before it replies. Here is what that means in everyday language, why it makes answers more trustworthy, and where it still goes wrong.
What is a context window in AI? A beginner's guide to how much a chatbot can "remember"
An AI chatbot can only read so much text at once before it starts forgetting the beginning. That limit is called the context window. Here is what it means in plain language, why long chats and big documents go sideways, and a few simple habits that keep the AI on track.
A glossary you navigate, not memorize.
Terms are learned by their distance from one another. We render the glossary as a network — click a node to open the term, every term links back into the map.