asking LLM's questions and seeing them cite sources and show their reasoning behind answers is way better than using google and forming an opinion off of the first few reddit answers you see, but you still have to look at what sources they're citing and make sure the entire thing isn't made up from the ground up.
I've seen long arguments where the AI is insisting it's correct just because a general consensus around one incorrect source and after pointing out that the source it was using is completely unreliable it started to shift it's opinion slightly. when it get's to the point where it's using logic and reasoning to spot bullshit sources and builds that into it's initial answers it will be even better.