Seven AI terms that keep coming up in every tool discussion: agents, RAG, vector databases and more

Agents and RAG are the two that matter most for understanding how current AI products work. The others, vector databases, ASI, fine-tuning, embeddings, hallucination, fill out the vocabulary needed to follow most technical conversations without losing the thread.

The reason shared vocabulary matters for this community specifically: when someone asks whether a tool supports RAG or whether an AI assistant is agentic or reactive, those are not casual uses of jargon. They are precise distinctions that change what the tool is capable of. Getting those distinctions clear makes the tool discussions in the other threads more useful.

Worth using this as the jumping-off point for a community glossary thread. Which AI terms still feel unclear or have definitions you keep forgetting? Happy to put together a living glossary post if there is enough appetite.