Klaudia Jaźwińska & Aisvarya Chandrasekar
Save this paper to a shelf, write a review, and keep your own notes.
main points: - ai hates to admit when they don't know something; gives incorrect/wrong answers confidently (premium chatbots provided more inaccurate answers) - fabrication of links/sources or use information from sources they weren't supposed to have to access to - even when using sources they had access to, they didn't cite more accurately