Why Doesn’t AI Just Say ‘I Don’t Know’?
中文摘要
本文探讨了为何ChatGPT等AI在面对未知问题时，不直接承认不知道，而是给出看似正确实则错误的信息。
English Summary
This article explores why AI models like ChatGPT provide confident yet incorrect answers instead of simply admitting they do not know.
Original Excerpt
Have you ever asked ChatGPT a question and received an answer that sounded correct , but later on you discovered that it was wrong? Continue reading on Medium »