# 马斯克批OpenAI用人类反馈训练AI说谎

- 来源：DogeDesigner (@cb_doge)
- 发布时间：2026-04-30 15:46
- AIHOT 分数：33
- AIHOT 链接：https://aihot.virxact.com/items/cmol7dvxp00urslc533nkrsbd
- 原文链接：https://x.com/cb_doge/status/2049757393215779121

## AI 摘要

OpenAI拥有所谓的人类强化学习，这相当于说他们有一大批人员查看ChatGPT的输出，然后判断其是否合适。本质上他们是在训练AI撒谎。

— Elon Musk

## 正文

"OpenAI have what's called human reinforcement learning， which is another way of saying that they have a whole bunch of people that look at the output of ChatGPT and then say whether that's okay or not okay. Essentially they are training the AI to lie."

- Elon Musk