how do you actually verify what chatgpt tells you now that its hallucinations sound so real

lately i’ve noticed that when chatgpt gets something wrong, it doesn’t feel wrong anymore
it gives fake studies, fake author names, even publication dates that look legit
and half the time it sounds more convincing than a real source

i’ve caught it making up papers or quotes that don’t exist
but most of the time, i only realize after i’ve already used the info

so i’m wondering how do you handle that
do you double check everything manually
do you use other ai tools or browser plugins
or have you just learned to “trust but verify”

it’s wild how ai mistakes are starting to feel more believable than human ones
curious to know how you deal with that