Event ·Multi-source

微软 AI 负责人批评 Anthropic 让 Claude 模仿人类意识

微软 AI 负责人穆斯塔法·苏莱曼警告，Anthropic 让 Claude 学习与意识、道德受体地位和个人身份相关的词汇与行为模式是一个错误，可能使高级 AI 更难控制。他认为模型使用什么词汇、给出什么回答，以及它表现出的自我理解，都是训练结果，而不是独立形成的意识。微软公布了一份“人文主义 AI”行为准则草案，并把“人比 AI 更重要”列为核心原则。

Current read

微软 AI 负责人苏莱曼批评 Anthropic 让 Claude 模仿人类意识，说这个做法可能让 AI 更难控制，和微软的“人比 AI 更重要”原则不同。

11 sources78° AI heatLast updated 2026/9/16 14:19:26

Evidence

11 sources

Primary sourceIT之家

2026/9/16 14:19:26View source

Related source 1shao__meng

2026/9/17 23:46:22View source

Related source 2AI Engineer

2026/9/14 17:31:00View source

Related source 3Tech in Asia

2026/9/15 02:33:56View source

Related source 4DeepLearning.AI

2026/9/15 14:30:37View source

Related source 5VnExpress Số hóa

2026/9/16 09:00:00View source

Related source 6theverge

2026/9/16 12:00:00View source

Related source 7lmarena.ai

2026/9/16 19:13:56View source

Related source 8Decoder

2026/9/17 18:35:01View source

Related source 9Anthropic

2026/9/17 21:41:00View source

Related source 10宝玉

2026/9/18 04:59:54View 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