What is retrieval-augmented generation (RAG)?
Retrieval-augmented generation (RAG) is a technique where an AI retrieves relevant information from an external source and uses it to generate a more accurate, up-to-date answer - instead of relying only on what the model memorized.
A language model on its own answers from what it learned during training, which can be stale or generic. RAG adds a retrieval step: the system finds relevant, current information - from the web, a knowledge base, or your own data - and feeds it to the model so the answer is grounded in real sources.
RAG is how modern AI search and assistants stay accurate and current. It also underpins generative search engines, which retrieve and cite sources when they answer.
how this works in folk
Folk uses retrieval to ground its answers in real, current information - browsing the live web on its own computer and pulling from your personal memory - so responses fit your actual context, not just the model's training.
frequently asked
Why is RAG useful?
It makes AI answers accurate and up to date by grounding them in retrieved sources, and lets an assistant use your own data or the live web instead of relying only on what the model memorized.
How does folk use retrieval?
Folk browses the live web on its own cloud computer and retrieves from your persistent memory, so its answers and actions reflect current information and your personal context.