[logseq-plugin-git:commit] 2024-09-28T14:11:54.411Z

journals/2024_09_28.md

@@ -1,7 +1,7 @@
 - 事实核查员机器人 #定投课堂 #ChatGPT
 	- 答案总是显得非常的斩钉截铁
 	- 反馈不一定靠谱
-		- 靠谱方面就是局限
+		- ChatGPT在靠谱方面就是局限性
 	- > What's the limitation of ChatGPT? Give me a comprehensive answer.
 	- > ChatGPT has several limitations:
 		- Accuracy: It can generate incorrect or outdated information. Its knowledge is based on data up to 2021, so it may not reflect recent events or developments.
@@ -53,7 +53,6 @@
 	- AI内容生成机制是依赖“概率运算”的。
 		- > Provide a list of prompts that are likely to trigger AI hallucinations, and explain why hallucinations would possibly happen.
 		  > 提供一个可能出发人工智能幻觉的提示列表，并解释为什么可能出现幻觉
-		- 我的 ChatGPT 聊天记录：
 		- https://chatgpt.com/share/66f39ela-d8d8-8009-b507-6b558bda92ff
 	- 事实核查员角色提示
 		- https://chatgpt.com/share/66f37a6f-620-8009-b284-36ac335bf5e