云上车间中的劳动折叠:人工智能数据标注的劳动过程研究

Folded Labor in Cloud Workshop:A Study on the Labor Process of Data Annotation Workers

数据标注员是当前人工智能产业从业者中的一个典型工种。通过对其具体劳动过程的研究发现,平台技术对劳动过程的一个显著改变在于部分劳动呈现为一种“幽灵化”的状态:首先,工厂车间的云端化成为幽灵劳动诞生的前提。以“公会”为核心形式的众包式、分包式网络可以完成从认领工作、准备生产材料到完成生产、分发薪资等一系列流程的组织。其次,可见性的改变是劳动幽灵化的核心轴承,平台技术通过对可见性的精巧控制,实现了劳动者与监督者肉身的实际消失。进而,劳动协作不再是人与人之间的流水线生产,而变成了机器安排下的技术延伸。最后,以平台为中心的数字关系本身也呈现为一种幽灵般的抽象结构,它遮蔽了传统的社会关系,使得劳动者的职业认同显得矛盾重重。

Data annotators are a typical occupation among practitioners in the AI industry.This article argues that one of the important changes in the labor process due to platform technology is that some types of labor have become“ghosted”.Firstly,the cloud workshops provide the foundation for the emergence of ghost labor.A network organized in the form of“guilds”through crowdsourcing and subcontracting can accomplish a series of processes,including job assignment,preparation of production materials,completion of production,and salary distribution.Secondly,changes in visibility are central to this ghost process,in which platform technology allows for the physical disappearance of workers and supervisors through its sophisticated control of visibility.In addition,the collaboration of labor no longer follows an assembly line production mode,but instead becomes a technical extension of the machine arrangement.Finally,the platform-centered digital relationship itself is presented as a ghostly abstract structure that obscures traditional social relations and creates contradictions in workers'professional identities.