Yes, which is why the ability to sift accurate and authoritative sources from spam, propaganda, and intentionally deceptive garbage, like advertising, and present those high-quality results to the user for review and consideration, is more important than any attempt to have an AI serve a single right answer. Google, unfortunately, abandoned this problem some time ago and is now left to serve up nonsense from the melange of low-quality noise they incentivized in pursuit of profits. If they had, instead, remained focused on the former problem, it’s actually conceivable to have an LLM work more successfully from this base of knowledge.