人机劳动分工:生成逻辑、模式类型与作用机制

Human-machine Labor Division:Generation Logic,Types and Mechanism

人机分工协作型混合劳动力已成为企业数智化转型的新生动能,但相关理论研究仍明显滞后于组织人机混合用工实践,且缺乏适配的系统理论体系。本文基于行动者网络理论与劳动分工理论,将智能机器视作与人并肩分工协作的劳动行动主体,从劳动分工视角创新性提出人机分工论作为诠释人机混合劳动力的理论新范式,并从生成逻辑、理论内涵、模式类型、作用机制等多维度探讨人机分工理论体系。结果发现:(1)人机分工作为第四次社会分工的重要体现,以劳动价值共创为基本目标,强调从传统以人为单一主体的一元分工转向人机双元分工,并可将人机分工结构划分为纯人工执行、机器赋能人工、人机协同执行、人工辅助机器、纯机器执行等模块。(2)机器对人工仅是劳动功能局部性超越,人工在人性化软技能、机器在程序化硬技能领域各占优势,且人机分工模式多元化可从主导地位、协作方式、数量匹配、管理对象等多重视角细分。(3)“机器换人”是一个让机器从幕后走到台前、人工从台前转向幕后的人机分工重构过程,其本质是分解工作内容并让人更像人(而非像机器)一样工作;同时,人机分工中“机器换人”与“机器增强”并行、悖论叠加渗透、工序拆分与功能拆解双向路径交互融合,且从时间、空间、功能等多维度呈现循环式人机分工动态演化选代。人机分工论可为诠释人机混合劳动力现象提供理论借鉴,也可为引领企业组织混合用工实践与重塑人机分工体系提供理论支撑。

The human-machine hybrid workforce(HMHW)has already become a new vitality in the enterprises'digital and intelligent transformation,while the relevant research currently lags far behind enterprise practice and lacks a systematic theoretical framework.Based on the Actor Network Theory(ANT)and the theory of labor division,this study regards intelligent machines as labor action subjects that work side by side with humans in the process of labor division.Meanwhile,this paper innovatively proposes the theory of human-machine labor division(HMLD)as a new theoretical paradigm for interpreting HMHW from the perspective of labor division,and discusses the theoretical system of HMLD from multiple dimensions such as generation logic,theoretical connotation,mode type,and action mechanism.The key findings of this study can be summarized as follows.First,as an important embodiment of the fourth social division of labor,HMLD takes labor value co-creation as its basic goal and emphasizes the shift from the traditional unitary labor division of human to dual action subject of HMLD,and the structure of HMLD can be divided into modules including pure manual execution,machine augmenting human,human-machine collaborative execution,human assisting machine,and pure machine execution.Second,both human beings and machines have their pros and cons,and machines transcend human beings only in part of their functions.And the modes of HMLD are diversified and can be subdivided from multiple perspectives,such as dominance,collaboration,quantity matching,and management object.From the perspective of the dominant role of human and machine,the types of HMLD can be categorized into five types,including fully manual,human-dominated labor division,human-machine collaborative-dominated labor division,machine-dominated labor division,and fully automated.From the perspective of cooperation mode,there are four types of HMLD,including human-machine separation,human-machine mutual substitution,human-machine interactive empowerment,and human-machine hybrid collaboration.From the perspective of quantity matching of human and machine,HMLD modes can be divided into multiple types,such as single human and single machine cooperation,single human and multiple machine cooperation,multiple human and single machine cooperation,and multiple human and multiple machine cooperation.From the dimension of management objects,the management modes of HMLD can be divided into five types,including human managing human,human managing machine,machine managing machine,machine managing human,and human-machine hybrid management mode.Third,the contrasting paradoxical effects of replacing human labor with machine(RHLM)and machine-augmentation in HMLD simultaneously exist and mutually influence each other.RHLM is a reconstruction process of HMLD,in which machines move from behind the scenes to the front stage,and humans turn from the front stage to behind the scenes.Essentially,the ultimate goal of RHLM is to decompose and refine work tasks,enabling machines to release human labor,and allowing humans to work more like real human beings instead of machines.On the other hand,machine-augmentation aims to empower human labor with machines,achieving complementary advantages between humans and machines and improving the efficiency of human-machine teams.Fourth,HMLD exhibits a cyclical and dynamic evolutionary iteration across multiple dimensions including time,space,and function.From the temporal perspective,HMLD exhibits a dynamic evolutionary trajectory from pure manual execution to human-dominated division of labor,followed by human-machine collaboration,machine-dominated division of labor,and finally,fully automated machine execution.From the spatial perspective,HMLD demonstrates a dynamic evolutionary pattern characterized by primarily internal circulation,supplemented by external circulation,with both internal and external circulation complementing each other.Meanwhile,employees replaced by machines exhibit a dual flow pattern of both internal and external circulation,and manifest in various directions,such as upgrading,downgrading,retaining current positions,transferring positions,natural attrition,employee sharing,and layoffs.From the functional perspective,HMLD exhibits a multi stage dynamic evolution from complete manual labor division to operational function division,power function division,and finally control function division.The conclusions of this paper can provide theoretical reference for interpreting the phenomenon of the symbiotic HMHW,and can also provide theoretical support for leading the practice of digital and intelligent transformation of enterprises and constructing a new labor division system.Compared with existing literature,the theoretical contributions of this study lie in the following two aspects.Firstly,the study extends the application of ANT to interpret organizational HMLD,treating intelligent machines as primary labor actors.This approach transcends the conventional labor division theory that regards humans as the sole actors.By innovatively proposing the theory of organizational HMLD with dual human-machine labor actors from the perspective of labor division,this study expands the traditional human-centered framework to encompass both humans and machines as dual labor actors.This theoretical advancement not only deepens and broadens the existing labor division theory,but also provides foundational theoretical support for interpreting human-machine collaboration in the digital era.Secondly,this research innovatively constructs and analyzes the basic theoretical framework of HMLD from multiple dimensions.It systematically examines the generative logic and theoretical implications of HMLD and categorizes the mode types,mechanisms,and typical practices of HMLD from various perspectives.This analysis offers insights for the further exploration and refinement of the HMLD theory,providing valuable theoretical references for the practice of HMHW in organizational contexts.Based on the above,this study proposes the following research agenda for future studies.Firstly,in-depth case studies of typical HMLD scenarios.For instance,by selecting typical enterprises adopting the HMHW,case study methods can be employed to further explore and enrich the basic theoretical framework of HMLD.Secondly,exploration of the influencing factors and driving mechanisms of HMLD.For example,grounded theory and qualitative meta-analysis can be applied to systematically construct an influencing factor model for HMLD.Moreover,methods like QCA can be used to explore the interactive mechanisms and configurational effects among various factors in complex HMLD systems.Thirdly,investigation of human-machine behaviors in HMLD systems.Issues such as individual psychological and behavioral changes of human employees,human-machine trust mechanisms,machine behavior,role positioning and hierarchy in HMLD,and all these areas require an in-depth theoretical exploration to construct a new paradigm of human-machine hybrid organizational behavior.