微软周一发布了一份针对AI训练的临时行为准则，旨在对其人工智能能力设置限制，以回应业界对AI技术失控的担忧。1该准则由微软AI首席执行官穆斯塔法·苏莱曼主导制定，耗时数个月完成。1
根据这份准则，微软的AI模型被禁止处理与武器开发、暴力或色情内容以及危险物品采购相关的请求。1苏莱曼表示，"AI必须从属于人类，始终服务于人类"，1这份指南仅是公众咨询的初步版本。1微软首席执行官萨提亚·纳德拉也强调，"如果我们开发的AI不是在帮助人类且处于人类控制之下，那就不值得追求"。1
微软的此举呼应了业界其他领导人最近提出的AI安全警告。1Anthropic首席执行官达里奥·阿莫代伊曾警告AI行业需要放慢脚步，1OpenAI首席执行官山姆·奥特曼表示，"我们欢迎一个为前沿AI设定一致性安全要求的联邦框架"。1
Microsoft released a preliminary code of conduct for AI training on Monday, aimed at restricting its artificial intelligence capabilities in response to growing industry concerns about uncontrolled technology development.1 The guidelines, developed over several months, prohibit AI models from processing requests related to weapon development, violent or sexually explicit content, and assistance with acquiring dangerous materials.1
Microsoft AI Chief Executive Mustafa Suleiman stated that "AI must be subordinate and always in service of people," emphasizing the company's commitment to responsible development.1 Satya Nadella reinforced this position by declaring: "If the AI we build is not helping humanity and under human control, it's not worth pursuing."1 The company described the code of conduct as an initial version open to public consultation.1
The announcement aligns with recent safety warnings from other AI industry leaders. Dario Amodei has called for the sector to "slow down," while Sam Altman of OpenAI stated: "We welcome a federal framework that sets consistent safety requirements for frontier AI."1
评论
还没有评论，欢迎留下第一条。