Ok, thanks for assuming so many things about me and for contributing to the discussion.

That's exactly the point that I'm making - they work really well for some tasks, extremely bad for other. People need to know how to use them in order for them to be useful. If one doesn't understand what an AI is it will only be a useless and potentially dangerous tool.

You can trust them, just for other tasks than this one. You're asking a specific tool to do something that it cannot do - ofc you can't trust the results. But if you use AI for stuff it can do, you'll see how powerful of a tool it actually is.

I think you're misunderstanding what an AI is and how it can help. It cannot do the specific task you asked it to do. It can however help on so many other tasks - but it needs to be trained and, most importantly, understood in order to be a useful helper.

Yes and no. AIs are stupid, but extremely knowledgable - more than a human ever will. For the specific task of factchecking there are bots created and trained to do that. Also, the paid version is way more accurate and can carry out web researches. But yeah everything needs human supervision.