Comment by imiric

How is LLM+RAG any different from a fine-tuned LLM without RAG? They're both trained on data, and have the capability to hallucinate.