# LLM 长任务语言漂移问题

- 来源：Ethan Mollick (@emollick)
- 发布时间：2026-09-21 01:52
- AIHOT 分数：32
- AIHOT 链接：https://aihot.news/items/cmua4f0xh0dunrowj4v7tyosr
- 原文链接：https://x.com/emollick/status/2101731295659180527

## AI 摘要

讽刺的是，如今用大语言模型跑长时间智能体任务时，最让人头疼的不是写代码、不是报错、也不是模型幻觉，而是随着任务推进，它们的语言会因漂移而变得越来越差

名字里就写着呢！写得好一点不行吗！

## 正文

It is ironic that the thing that is now most annoying about long-running agentic tasks with Large Language Models isn't coding or errors or hallucinations, but the fact that their language gets worse due to drift as a task goes on

Its in your the name! Just write better already!