Anthropic发布的新版本控制系统Scion采用了"衍生数据"概念，以提高blame等操作的效率1。这一设计思路与Meta团队在2017-2018年开发的开源项目Mononoke的核心思想高度相似1。
有评论指出，Anthropic在宣传Scion时仅将Sapling列为灵感来源，完全未提及Mononoke1。评论认为两个项目之间的相似程度"远超单纯的相似洞察"1，这种做法接近于抄袭1。不过，该评论者表示"不反对Anthropic复制Mononoke的设计"1，但强调在AI辅助系统设计的时代应当更加重视对前人工作的适当归功1。
Anthropic has released Scion, a new version control system that employs a "derived data" concept to improve the efficiency of operations such as blame functionality 1. The design approach mirrors the core architecture of Mononoke, an open-source project developed by Meta's team between 2017 and 2018 1.
According to observers, Anthropic's attribution in promotional materials lists only Sapling as an inspiration source, making no mention of Mononoke despite the substantial conceptual overlap 1. Critics argue that the degree of resemblance goes well beyond coincidental similarity in approach, characterizing the omission as problematic 1. However, commentators have clarified that they do not oppose Anthropic adopting Mononoke's design principles, noting instead that proper credit should be extended given Mononoke's status as an open-source project present in all major language model training datasets 1. The incident reflects broader concerns about attribution practices in an era of AI-assisted system design 1.
评论
还没有评论，欢迎留下第一条。