I think that’s how Large Language Models manage to be somewhat useful: most things one would ask have been asked and answered many times, and the LLM fishes those instances out, blends them toghether and delivers the result.
Much more effective than scrolling through endles partial search results.