Does Artificial Intelligence Challenge the Labor Theory of Value?

In recent years, the latest developments in artificial intelligence are profoundly changing people’s ways of life and production. In this process, a thought or doubt follows like a shadow. That is, with the development of computer technology and artificial intelligence, is the Marxist labor theory of value still valid? The puzzle of “machines replacing humans” has once again been placed before people in a new form. Relatedly, how to view so-called “immaterial labor” and “digital labor,” and their relationship with material production, which is the core of historical materialism, has also become a focal topic of recent academic discussion. The viewpoint of this article is: the development of artificial intelligence does not constitute a challenge to the Marxist labor theory of value; in the era of the digital economy, the Marxist labor theory of value has not become obsolete; it constitutes the scientific tool for us to analyze the social-historical effects of artificial intelligence. Overcoming the anxiety that artificial intelligence challenges the labor theory of value requires dispelling the myths surrounding artificial intelligence and clarifying misunderstandings about the Marxist labor theory of value.

近年来，人工智能的最新发展正在深刻改变人们的生活和生产方式。在此过程中，一种思考或疑惑如影随形。这就是伴随计算机技术和人工智能的发展，马克思主义的劳动价值论还有效吗？“机器替代人”的难题以新的形式又一次摆在了人们的面前。与之相关，如何看待所谓的“非物质劳动”“数字劳动”，及其与作为唯物史观核心的物质生产之间的关系，也成为近期学界关注的焦点话题。本文的观点是：人工智能的发展并不能构成对马克思主义劳动价值论的挑战；在数字经济时代，马克思主义劳动价值论并未过时，它构成了我们分析人工智能社会历史效应的科学工具。克服人工智能挑战劳动价值论的忧思，需要破除围绕人工智能的迷思，澄清有关马克思主义劳动价值论的误解。

Grasping the Labor Theory of Value with Historical Materialism

以唯物史观把握劳动价值论

In order to clarify that the development of artificial intelligence does not constitute a challenge to the Marxist labor theory of value, we might first conduct a conceptual analysis of the labor theory of value. Although we often say that Marx expounded a scientific labor theory of value in *Capital*, returning to Marx’s classic works, it is difficult to find Marx’s direct use of the term “labor theory of value.” Using the labor theory of value to describe Marx’s theory of value was gradually established after the second half of the 19th century, following the development of Western economics, especially after the marginal revolution, in opposition to the “utility theory of value.” [1]1 The Luddite movement refers to early 19th-century English textile workers who destroyed machinery they blamed for unemployment and wage cuts, often used as a symbol of resistance to technological change. Both Western economics and Marxist political economy emphasize the close connection between Marx and Ricardo, or the classical political economists, on the issue of the labor theory of value. However, as represented by Samuelson, in the eyes of Western economists, Marx’s labor theory of value is seen as a further development of economics along Ricardo’s erroneous path. [2]2 Paul Samuelson (1915–2009) was a prominent American economist and Nobel laureate, known for his neoclassical synthesis. Within the theoretical framework of Marxist political economy, Marx’s labor theory of value thoroughly criticized the labor theory of value of classical political economy and constituted the scientific premise of the theory of surplus value. So, how exactly did Marx expound the labor theory of value?

为了阐明人工智能的发展并不能构成对马克思主义劳动价值论的挑战，我们不妨首先对劳动价值论做一个概念辨析。尽管我们经常谈到马克思在《资本论》中阐发了一种科学的劳动价值论，但是回到马克思的经典著作中，我们很难找到马克思对“劳动价值论”这个术语的直接使用。用劳动价值论来描述马克思的价值理论，是19世纪下半叶之后随着西方经济学的发展，特别是边际革命之后，在同“效用价值论”相对立的意义上逐渐确立起来的。无论是西方经济学还是马克思主义政治经济学，都强调马克思与李嘉图，或者古典政治经济学家，在劳动价值论问题上的密切关联。只不过，以萨缪尔森为代表，在西方经济学家眼中，马克思的劳动价值论被看作经济学演进在李嘉图的错误路向上的进一步发展。而在马克思主义政治经济学的理论框架中，马克思的劳动价值论彻底批判了古典政治经济学的劳动价值论，构成了剩余价值理论的科学前提。那么，马克思到底是怎样阐述劳动价值论的呢？

In everyday conceptions, when we mention Marx’s labor theory of value, or demonstrate the scientific nature of the Marxist labor theory of value, we often proceed from the following two aspects. First, only human labor can create value; therefore, when discussing the value of commodities, one can only start from labor. Second, the reason Western economics denies the labor theory of value is its inability to understand the objective abstractness of socially necessary labor as value. Strictly speaking, both of these lines of thought are incomplete. Returning to the textual structure of *Capital* and its creative process, we can see: when Marx wrote the final draft of Volume I of *Capital*, he rewrote the chapter discussing commodities and made the firm decision to use the category of “value.” In this part, “use-value” and “value” constitute the dual attributes of the commodity, behind which lies the dual character of labor. That is to say, the reason Marx critically transformed the labor theory of value of classical political economy is itself directly related to the understanding of the dual character of labor.

在日常的观念中，当我们提到马克思的劳动价值论，或者论证马克思主义劳动价值论的科学性时，往往从这样两个方面展开。其一，只有人的劳动才能创造价值，因此讨论商品价值时只能从劳动出发。其二，西方经济学之所以否定劳动价值论，是因为无法理解作为价值的社会必要劳动的客观抽象性。严格来说，这两种思路都是不全面的。回到《资本论》的文本结构及其创作过程中，我们可以看到：马克思在写作《资本论》第一卷最后一稿的时候，重写了论述商品的一章，并下定决心使用“价值”这个范畴。在这一部分，“使用价值”和“价值”构成了商品的二重属性，背后则是劳动的二重性。也就是说，马克思之所以批判地改造古典政治经济学的劳动价值论，本身是同对劳动二重性的理解直接相关的。

To correctly understand Marx’s labor theory of value, one must adhere to historical materialism as a scientific conception of history. The distinction of the dual character of labor, and the discovery of social labor or abstract labor as the “substance” of commodity value, can only be obtained within the horizon of historical materialism. Only by adhering to the starting point of the mode of material production can one truly understand the overall relationship of the mode of production that exists behind the exchange relations of various commodities. It is precisely from this perspective that we can truly understand what Marx meant by the “value” of a commodity, and its distinction from “wealth” and “exchange-value.”

正确理解马克思的劳动价值论，必须坚持唯物史观这一科学的历史观。劳动二重性的区分，对于作为商品价值的“实体”的社会劳动或抽象劳动的发现，只能在唯物史观的视域中获得。只有坚持从物质生产方式出发，才能真正理解各种商品的交换关系背后存在着的生产方式的总体性关系。也正是从这一视角出发，我们才能真正理解，什么是马克思所说的商品的“价值”，及其与“财富”“交换价值”之间的区别。

The “value” understood by Marx is, in essence, a certain proportional relationship of commodity exchange at a specific social-historical stage. This proportional relationship is ultimately determined by a specific mode of production. “Value” manifests as the “exchange-value” of a commodity in concrete exchange relations. This “exchange-value,” expressed in monetary form, is “price.” The “wealth” often mentioned in everyday contexts, on the one hand, combines with our perception of “use-value,” acquiring the meaning of material wealth; on the other hand, for classical political economists, especially Adam Smith, it also functions, in a fetishistic form, to justify capital accumulation.

马克思所理解的“价值”，在本质上是在一定的社会历史阶段，商品交换的一定的比例关系。这种比例关系归根结底是由一定的生产方式所决定的。“价值”在具体的交换关系中表现为商品的“交换价值”。这种“交换价值”以货币形式来表现就是“价格”。日常语境中经常提到的“财富”，一方面与我们对“使用价值”的感受结合在一起，具有了物质财富的含义；另一方面，在古典政治经济学家，特别是亚当·斯密那里，也以拜物教的形式承担着论证资本积累的功能。

Going further, precisely because Marx’s theory of value itself must be understood by returning to the material production process, commodity exchange relations must also be understood as the manifestation of relations of production. Thus, for the labor theory of value, it inherently contains two theoretical dimensions. First, the labor theory of value does not simply emphasize that labor creates value, or that labor constitutes the “substance” of value; rather, it must be based on the analysis of the mode of material production, focusing on the mode of constitution of actual social relations. Second, this social relation is not an isolated, abstract relationship between two commodity owners, but a social-historical totality unfolded through commodity relations. Within such a social-historical totality, the connections between things are real, objective, and universal, but because of the existence of various mediations, the connections between things often present themselves to people in different forms.

更进一步，正是因为马克思的价值理论本身必须回到物质生产过程中获得理解，商品交换关系也就必须被理解为生产关系的表现形式。这样，对于劳动价值论来说，内在蕴含着两个方面的理论维度。其一，劳动价值论不是简单强调劳动创造价值，或是劳动构成了价值的“实体”，而是必须立足物质生产方式的分析，关注现实的社会关系构成方式。其二，这种社会关系不是孤立的、抽象的两个商品所有者之间的关系，而是一个以商品关系展开的社会历史的总体性存在。在这样一个社会历史的总体性存在中，事物之间的联系是现实的、客观的、普遍的，但因为各种中介的存在，事物的联系往往以不同的形式呈现在人们面前。

The fundamental reason Marx’s labor theory of value differs from the labor theory of value of classical political economy and the value theory of Western economics is that it starts from the mode of material production and scientifically elucidates the essence and manifestations of actual relations of production and social relations. This lays a solid theoretical foundation for us to correctly understand the development of the digital economy, grasp the social-historical effects of the development of artificial intelligence, and respond to the so-called theoretical challenges posed by artificial intelligence.

马克思的劳动价值论之所以从根本上不同于古典政治经济学的劳动价值论，以及西方经济学的价值理论，就在于它从物质生产方式出发，科学阐明了现实的生产关系、社会关系的本质及其表现形式。这为我们正确认识数字经济的发展，把握人工智能发展的社会历史效应，回应人工智能提出的所谓理论挑战奠定了坚实的理论基础。

Analyzing Labor in the Age of Artificial Intelligence with the Labor Theory of Value

以劳动价值论解析人工智能时代的劳动

Since the beginning of the 21st century, global scientific and technological innovation has entered an unprecedentedly intensive and active period. A new round of technological revolution and industrial transformation is reshaping the global innovation landscape and restructuring the global economic structure. In the past decade, with the rise of the “Fourth Industrial Revolution,” the development of the internet and big data, and the emergence of various cutting-edge and disruptive technologies, artificial intelligence has once again attracted people’s attention. [3]3 The “Fourth Industrial Revolution” is a term popularized by Klaus Schwab, founder of the World Economic Forum, referring to a current era of technological change characterized by the fusion of digital, physical, and biological domains. The reason for saying “once again” is that the concept of “artificial intelligence” was proposed as early as 1956. However, today’s development of AI technology, whether in terms of data and computing power, or in its applications and impact, far surpasses the past, leading many to harbor such anxieties: the development of AI has already produced new forms of control over people by data, algorithms, and platforms. AI has already replaced humans in many fields. In the future, will AI develop autonomous consciousness and become the master of humanity? We should respond to these questions separately.

进入21世纪以来，全球科技创新进入空前密集活跃的时期，新一轮科技革命和产业变革正在重构全球创新版图，重塑全球经济结构。近十年来，伴随“第四次工业革命”的兴起、互联网和大数据的发展、各种前沿性颠覆性技术的出现，人工智能重新受到人们的关注。之所以说“重新”，是因为“人工智能”这个概念早在1956年便已经被提出，只不过今天人工智能技术的发展，无论在数据和算力，还是在应用和影响上都远超以往，以至于很多人都会产生这样的忧虑：人工智能的发展，已经产生了数据、算法、平台对人的新的控制。人工智能在很多领域都已经替代了人。未来，人工智能会不会产生自主意识，成为人类的主宰呢？对于这些问题，我们应该分别加以回应。

First, regarding the current state of AI technology development, computer scientists and AI researchers are still primarily conducting research around weak AI, that is, better utilizing computer technology and internet development to solve problems encountered in real life. The question of whether machines can have autonomous consciousness, or the issue of strong AI, is not a direct concern of scientific and technological development. In current discussions, this issue is closer to science fiction than to science.

首先，就目前人工智能技术的发展来说，计算机科学家和人工智能研究者仍然主要是围绕弱人工智能展开研究的，也就是更好地利用计算机技术和互联网的发展，来解决现实生活中遇到的问题。所谓机器能否有自主意识的问题，或者说强人工智能问题，并不是科学技术发展直接关注的问题。在现阶段的讨论中，这一问题更接近于科幻而非科学。

Second, in actual social life, does control over people by data, algorithms, and platforms exist? Objectively speaking, along with the development of AI, such phenomena do indeed exist, such as platforms capturing users’ personal information, behavioral habits, and even privacy, and the control and discipline of people by data and algorithms in production activities. However, these phenomena of so-called “data controlling people” are, in essence, merely a reprint of the problem of “abstract forces enslaving people” that Marx had already confronted and analyzed.

其次，在现实的社会生活中，是否存在数据、算法、平台对人的控制呢？客观来说，伴随人工智能的发展确实存在这样一些现象，如平台对用户个人信息、行为习惯甚至隐私的抓取，数据和算法在生产活动中对人的控制和规约等。但是，这些所谓的“数据控制人”的现象，从本质来说不过是马克思已经面对并分析过的“抽象力量奴役人”的问题的再版罢了。

Third, regarding the issue of artificial intelligence replacing humans, objectively speaking, such a trend does indeed exist. As early as 2013, foreign research had already pointed out that nearly half of employment positions in the United States face the risk of being replaced by technological innovation. In reality, certain tasks and labor that originally required direct human involvement can now be completed by artificial intelligence. However, the problem of "machines replacing humans" did not emerge only with the development of artificial intelligence; since the Industrial Revolution, this issue has been raised and considered in different forms.

再次，关于人工智能替代人的问题，客观来说，确实存在这样的趋势。早在2013年，国外研究便已经指出，美国有将近一半的就业岗位存在被技术创新所取代的风险。现实中，某些原本需要由人直接从事的工作和劳动，现在已经可以由人工智能来完成了。然而，“机器替代人”的难题并不是伴随人工智能的发展才出现的，从工业革命开始，这一问题便以不同的形式被提出并思考了。

A simple conclusion is: historical facts have fully demonstrated that the resolution of the two problems of artificial intelligence controlling and replacing humans cannot be accomplished simply through the Luddite movement of "smashing machines," but must involve analyzing those abstract forces that actually exist in social life and studying those concrete changes that actually occur in the production process. When we confront the development of artificial intelligence with such an attitude, Marx's labor theory of value will not only not be challenged by artificial intelligence, but will become a scientific tool for us to grasp the development of artificial intelligence and to study and judge its social-historical effects. The author's view is as follows.

一个简单的结论是：历史的事实已经充分证明，对于人工智能控制、替代人这两个问题的解决，不能简单地通过“砸碎机器”的卢德运动来完成，而必须剖析社会生活中现实存在的那些抽象力量，研究生产过程中现实发生的那些具体变化。当我们以这样的态度来面对人工智能的发展时，马克思的劳动价值论不仅不会受到人工智能的挑战，而且会成为我们把握人工智能发展、研判其社会历史效应的科学工具。笔者的观点如下。

First, artificial intelligence is in essence a machine or a tool, albeit a complex machine and an efficient tool. Today, the development of artificial intelligence technology benefits first and foremost from the development of the internet, especially the emergence of the Internet of Things. Among these, the most critical technological elements include the substantial enhancement of computer storage and computing capabilities, the rapid development and widespread application of detector technology, and the popularization of high-speed internet, among others. Without the preparation of these hardware technological conditions, artificial intelligence could not be realized even with algorithms, let alone applied. Thus, the problem of labor under conditions of artificial intelligence is transformed into a problem of the transformation of material modes of production and the transformation of human labor processes within the new round of technological revolution and industrial transformation. History and reality prove that Marx's labor theory of value has the most to say on this issue.

第一，人工智能在本质上是机器或者工具，只不过是一种复杂的机器、高效的工具。今天，人工智能技术的发展，首先得益于互联网的发展，特别是物联网的出现。其中，最为关键的技术要素包括计算机存储运算能力的大幅提升、探测器技术的高速发展和广泛运用，以及高速互联网的普及化等。没有这些硬件技术条件的准备，人工智能即便有算法也无法实现，更不用说运用了。这样，人工智能条件下的劳动问题，就转化为新一轮科技革命和产业变革中的物质生产方式变革和人的劳动过程转型的问题了。历史和现实证明：马克思的劳动价值论对于这一问题最有发言权。

Second, as we have mentioned, adhering to Marx's labor theory of value means adhering to the historical materialism that proceeds from the material mode of production. Faced with the development of artificial intelligence, we should neither, from the simple viewpoint that labor creates value, marvel at the phenomenon of artificial intelligence replacing humans and creating wealth, nor, from the viewpoint that abstract labor constitutes the value of commodities, disregard the economic and social changes caused by the development of artificial intelligence technology. To adhere to Marxist political economy, we must delve into the changes in the production process and labor forms generated by the development of artificial intelligence technology, and explore the laws of the operation of the modern social economy. In this process, we should be wary of those practices that abstractly label the various forms of labor involved in the development of artificial intelligence as "immaterial labor," "digital labor," or similar tags. This is because the labor of writing code to develop software is labor, the labor of designing blueprints to research and develop machines is also labor, and the labor of using artificial intelligence to carry out production activities is still labor. Apart from differences in objects, tools, and methods at the level of the production of specific products, there are also abstract quantitative differences in overall economic life, especially within commodity production and exchange relations. This difference can only be measured in terms of socially necessary labor time, except that the formation of socially necessary labor time exists not only within the same industry sector but also across different industry sectors. It is precisely in this sense that we can more deeply understand the positive driving role of technological innovation in the development of productive forces and social progress.

第二，正如我们提到的，坚持马克思的劳动价值论，就是坚持从物质生产方式出发的唯物史观。面对人工智能的发展，我们既不能从简单的劳动创造价值的观点出发，惊叹于人工智能替代人、创造财富的现象，也不应从抽象劳动构成商品价值的观点出发，漠视人工智能技术发展所导致的经济社会变革。坚持马克思主义政治经济学，必须深入人工智能技术发展所产生的生产过程和劳动形式变化中，探寻现代社会经济运行规律。在此过程中，我们应当警惕那些将人工智能发展过程中所关涉到的各种劳动形式，抽象地贴上“非物质劳动”“数字劳动”之类标签的做法。这是因为，写代码开发软件的劳动是劳动，设计图纸研发机器的劳动也是劳动，运用人工智能开展生产活动的劳动还是劳动。除了在具体产品的生产的层面存在对象、工具和方式的差别外，还在总体的经济生活中，特别是商品生产和交换关系中存在抽象的量的差别。这一差别只能在社会必要劳动时间的意义上加以衡量，只不过社会必要劳动时间的形成不仅存在于同一行业部门内部，而且存在于不同行业部门之中。也正是在这个意义上，我们才能更加深刻地理解科技创新对于生产力发展和社会进步的积极推动作用。

Third, just as Marx's labor theory of value does not simply discuss the exchange value of commodities, nor abstractly discuss the accumulation of wealth, so too, when using Marxist political economy today to examine those so-called new forms of labor that appear under conditions of artificial intelligence, we should distinguish the different dimensions inherent in the same object. For example, the relationship between mental labor and manual labor that Marx once discussed can be understood more clearly precisely under conditions of artificial intelligence. Mental labor and manual labor are not completely separate; today, the two are increasingly and tightly intertwined. The crux of the issue is not their distinction and different functions, but the interrelationship of different forms of labor within a given mode of production and production process. For instance, regarding the so-called "productive consumption" or "playbour" phenomenon, the crux of the issue is not that personal consumption simultaneously involves the production of data, or that the process of playing a game is also a labor process, but rather whether and how, under new technological conditions, these activities that originally belonged to personal life and were not included within the scope of productive labor are embedded into social-economic relations in a new way. Another example: when we discuss labor under conditions of artificial intelligence, we often pay excessive attention to the process of AI research and development and the impact of AI technology products on social life. However, what cannot be ignored is that the labor involved in preparing the hardware conditions for the development of AI technology, the large amount of simple, primary labor involved in data sorting and annotation, and the labor of using AI technology, especially network platforms, to carry out production activities, all also exist objectively and realistically. The analysis of the nature and function of these forms of labor is equally inseparable from the actual material relations of production and specific social-economic relations.

第三，正如马克思的劳动价值论不是简单地讨论商品交换价值，也不是抽象地谈论财富积累一样，今天运用马克思主义政治经济学看待人工智能条件下出现的那些所谓新的劳动形式的时候，也应当区分同一对象自身所蕴含的不同维度。比如，马克思曾经讨论过的脑力劳动和体力劳动的关系问题，恰恰可以在人工智能条件下得到更加清晰的理解。脑力劳动和体力劳动并不是截然分开的，今天二者越来越紧密地纠缠在一起。问题的关键不是二者的区分和不同作用，而是一定生产方式和生产过程中不同劳动形式的相互关系。比如，所谓的“生产性消费”或者“玩工”现象，问题的关键并不在于个人消费的同时也进行了数据的生产，或是玩游戏的过程也是劳动过程，而是新的技术条件下，这些原本属于个人生活、未被纳入生产性劳动范围的活动，是否、怎样被以一种新的方式嵌入社会经济关系之中了。又如，当我们讨论人工智能条件下的劳动时，往往过度关注人工智能研发的过程，以及人工智能技术产品对社会生活的影响。但不容忽视的是，人工智能技术发展硬件条件准备过程中的劳动，数据整理和标注过程中大量的简单初级劳动，利用人工智能技术特别是网络平台进行生产活动的劳动，也都客观现实地存在着。对这些劳动的性质和作用的分析，同样离不开现实的物质生产关系和一定的社会经济关系。

Today, the development of artificial intelligence is profoundly changing people's forms of labor, production processes, and social relations; this is an objective historical trend. But artificial intelligence is not a "myth"; it is a complex, systematic, and efficient tool, or rather a system of machinery. Marx's labor theory of value does not proceed from intuitive anthropological premises or abstract conceptual speculation to understand labor and value, but rather adheres to historical materialism and employs dialectics in its rational form as a scientific tool for grasping actual modes of production and social formations. In this sense, rather than agonizing over artificial intelligence's replacement of or control over humans, we should more actively explore how to better develop the digital economy and form new quality productive forces. In this process, Marxist political economy will inevitably further demonstrate its own theoretical charm.

今天，人工智能的发展深刻改变着人们的劳动形式、生产过程、社会关系，这是一个客观的历史趋势。但是人工智能不是一个“神话”，而是一个复杂、系统、高效的工具，或者说一种机器体系。马克思的劳动价值论，并不是从直观的人类学前提或是抽象的概念思辨出发来理解劳动和价值，而是坚持唯物史观，运用合理形态的辩证法，把握现实的生产方式和社会形态的科学工具。在这个意义上，我们与其纠结人工智能对人的替代、控制，不如更加积极地去探索如何更好地发展数字经济，形成新质生产力。在此过程中，马克思主义政治经济学必将进一步展现自身的理论魅力。

(The author is a professor at the Center for Studies of Marxist Social Theory, Nanjing University)

（作者系南京大学马克思主义社会理论研究中心教授）

Online editor: Huihui

网络编辑：慧慧