If less hallucinations are the goal, surely this is a bit over the top? Surely if you have the ground truth facts available, then fine-tuning for EVERY subject

If less hallucinations are the goal, surely this is a bit over the top? Surely if you have the ground truth facts available, then fine-tuning for EVERY subject area seems much more work than using the facts with retrieval augmented generation, and making sure that the facts line up?