This is easily proven incorrect. Just go to ChatGPT and say something incorrect and ask it to verify. Why do people still believe this type of thing?
And yet models get things wrong all the time, too.
And yet models get things wrong all the time, too.