Event ·Multi-source

Anthropic 工程师解释 Claude 为什么越聪明写作反而越差

Anthropic 工程师 Jackson Kernion 表示，新 Claude 模型主要针对数学、代码和面向 AI 的技术解释做优化，导致文风像信息密集堆砌，人类读起来生硬。Opus 5.5 尝试修复这一写作问题，但他认为纯写作能力上 Opus 4.6 仍然无可替代。

Current read

Anthropic 工程师亲口承认 Claude 新模型写得越来越怪，原因是训练偏重代码和数学，想写文案的可以看看 Opus 4.6 还是不是最优解。

11 sources73° AI heatLast updated 2026/9/23 15:14:34

Evidence

11 sources

Primary sourceDecoder

2026/9/23 15:14:34View source

Related source 1Akshay Kothari

2026/9/23 14:26:32View source

Related source 2theverge

2026/9/22 16:30:00View source

Related source 3Claude

2026/9/22 16:31:01View source

Related source 4Anthropic

2026/9/22 16:31:47View source

Related source 5cat

2026/9/22 16:39:47View source

Related source 6Alex Albert

2026/9/22 16:39:59View source

Related source 7歸藏(guizang.ai)

2026/9/22 16:41:08View source

Related source 8宝玉

2026/9/22 16:41:36View source

Related source 9OpenRouter

2026/9/22 16:44:34View source

Related source 10The Rundown AI

2026/9/22 16:48:01View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story