My direct experience mirrors this. LLMs can be useful as a tool and can save time. Also while I've found some prompting techniques to can help with accuracy, I've not seen a method of getting reliably accurate answers. It's critical to always double check all of the output.

Also, Often the stories are AI generated fake news stories with AI generated images. Shameless garbage.

Good point! I'm loving this platform, but need to intentionally spend more time here than the other noisier and spammier sites.

Isn't this line in just about every episode of 20/20 as well?