Social Contract

The idea of a social contract is often introduced as a philosophical abstraction: an imagined agreement through which individuals consent to authority in exchange for security and order. Yet historically, social contracts are not ideals invented first and implemented later. They are ex post equilibria—practical arrangements that stabilize societies after major shifts in how value is created and distributed. Whenever a new production technology fundamentally alters this balance, the prevailing social contract comes under strain and is eventually redesigned.

Artificial intelligence represents such a moment.

To understand why AI forces a renegotiation of the social contract, it is useful to examine how social contracts have changed in the past. Across history, large-scale technological transformations—agriculture, industrialization, mass production, and digitalization—have repeatedly broken the alignment between value creation, value capture, and legitimacy. When that alignment fails, societies experience instability until new arrangements emerge.

What happned in history?

In early agrarian societies, the shift from hunter-gatherer economies to settled agriculture enabled the accumulation of surplus and the concentration of land ownership. Egalitarian social structures became untenable. New hierarchies formed, justified by religious authority or military protection, and redistribution took the form of granaries, tribute, and ritual obligation. The social contract changed not because thinkers demanded it, but because the economic base no longer supported the old order.

A similar rupture occurred during the Industrial Revolution. Pre-industrial urban economies were governed by guild-based social contracts that restricted competition, protected skilled labor, and embedded economic life within moral and social norms. Mechanization destroyed the scarcity on which these arrangements depended. Skilled artisans were displaced by machines operated by unskilled labor, while capital replaced apprenticeship as the gatekeeper to production. The resulting social dislocation—urban poverty, labor unrest, and political instability—persisted for decades until new contracts emerged in the form of labor rights, unions, and eventually welfare states.

The twentieth century saw another stabilization as mass production, national markets, and strong state capacity produced a labor-centered social contract. Employment became the primary channel through which individuals accessed income, dignity, and social protection. If one worked, one could reasonably expect a stable livelihood. This arrangement underpinned both political legitimacy and social cohesion in much of the industrialized world.

Digitalization and globalization began to erode this contract by weakening labor’s bargaining power and shifting value toward capital and intellectual property. Still, human labor—particularly skilled labor—remained central to value creation. Education and reskilling were promoted as universal solutions to technological disruption.

Artificial intelligence breaks this assumption.

What is the social contract in AI era?

Unlike earlier forms of automation that primarily replaced human muscle or routine tasks, AI substitutes for judgment, coordination, and cognitive labor. More importantly, it enables economic output to scale with diminishing reliance on human input. Productivity becomes increasingly decoupled from employment. This structural shift undermines the core premise of the modern social contract: that labor is the primary source of value and the legitimate basis for income distribution.

As a result, familiar pillars of legitimacy begin to weaken. Wages no longer track contribution. Education no longer guarantees relevance. Employment no longer ensures social inclusion. When these links erode, inequality becomes structural rather than cyclical, and existing redistribution mechanisms lose effectiveness.

This dynamic is not limited to society at large; it appears at the organizational level as well. Firms, like societies, operate under implicit social contracts that define who creates value, who captures rewards, and why authority structures are justified. AI reshapes internal production functions by amplifying individual leverage, reducing coordination costs, and making expertise more replicable. Yet compensation, promotion, and decision-making structures often remain anchored in older models tied to headcount, hierarchy, or tenure. When value creation and value capture diverge inside organizations, legitimacy erodes, politics intensifies, and high performers disengage or exit.

History suggests that such misalignments do not resolve themselves through narrative alone. Social contracts are ultimately rewritten through structural change. In past transitions, this has meant redefining what counts as contribution, altering how surplus is distributed, and developing new legitimacy narratives that reflect material reality. In the AI era, this likely entails partial decoupling of income from employment, a shift from labor-based taxation toward capital and rent-based systems, and recognition of AI systems as collective infrastructure rather than purely private assets.

Crucially, social contracts are rarely redesigned proactively. More often, they are renegotiated under pressure, after periods of instability and conflict. The lesson of history is not that disruption can be avoided, but that its costs depend heavily on timing. Societies and organizations that adapt early gain resilience; those that delay are forced into reactive reforms under far less favorable conditions.

Artificial intelligence does not merely change how work is done. It challenges what work means as the foundation of economic life. The labor-based social contract that defined the modern era is already under strain. Whether its successor emerges through deliberate design or through crisis remains an open—and urgent—question.

社会契约

社会契约这一概念，往往被作为一种哲学抽象来介绍：一种想象中的协议，个人通过让渡部分自由，换取秩序与安全。然而从历史上看，社会契约并不是先被设计出来、再被实施的理想蓝图。它们更像是一种事后形成的均衡——在价值创造与分配方式发生重大变化之后，为了稳定社会而逐渐形成的现实安排。每当新的生产技术从根本上改变价值如何被创造、如何被分配，既有的社会契约就会承压，并最终被重塑。

人工智能，正是这样一个时刻。

要理解为什么 AI 必然迫使社会契约重新谈判，有必要回顾社会契约在历史中的演变。纵观人类历史，农业、工业化、大规模生产以及数字化等技术跃迁，一次又一次打破了价值创造、价值获取与政治合法性之间的既有对齐关系。当这种对齐失效，社会便会陷入动荡，直到新的安排出现。

在早期农业社会中，从狩猎采集向定居农业的转变，使得剩余得以积累，土地得以集中。原本相对平等的社会结构不再可持续。新的等级体系随之形成，并通过宗教权威或军事保护来加以正当化；再分配的形式则体现为粮仓、贡赋和仪式义务。社会契约的变化，并非源于思想家的呼吁，而是因为经济基础已无法支撑旧秩序。

工业革命时期也发生了类似的断裂。工业化之前的城市经济，运行在以行会为核心的社会契约之上：限制竞争、保护熟练劳动，并将经济活动嵌入道德与社会规范之中。机械化摧毁了这些安排赖以存在的“稀缺性”。熟练工匠被机器和非熟练劳动力取代，资本取代学徒制成为进入生产体系的门槛。由此产生的社会失序——城市贫困、劳工冲突与政治不稳定——持续了数十年，直到新的社会契约逐渐形成：劳工权利、工会，以及最终的福利国家。

二十世纪，随着大规模生产、全国性市场与强国家能力的确立，一种以劳动为中心的社会契约得以稳定下来。就业成为个人获得收入、尊严与社会保障的核心渠道。只要工作，便可以合理地期待一个稳定的生活。这一安排构成了现代社会政治合法性与社会凝聚力的重要基础。

数字化与全球化开始侵蚀这一契约：削弱劳动的议价能力，将价值更多地推向资本与知识产权。但即便如此，人类劳动——尤其是高技能劳动——仍然是价值创造的核心。教育与再培训被视为应对技术冲击的普适解法。

而人工智能，打破了这一前提。

与主要替代体力或重复性任务的早期自动化不同，AI 替代的是判断、协调与认知劳动。更关键的是，它使经济产出能够在越来越少依赖人类投入的情况下持续扩张。生产率开始与就业脱钩。这一结构性变化，直接动摇了现代社会契约的核心假设：劳动是价值的主要来源，也是收入分配合法性的基础。

随之而来的，是一系列合法性支柱的松动。工资不再反映真实贡献；教育不再保证持续相关性；就业也不再必然意味着社会融入。当这些连接被削弱，不平等便从周期性问题转变为结构性问题，而既有的再分配机制也逐渐失效。

这种动态不仅存在于社会层面，在组织内部同样清晰可见。企业与社会一样，也运行在隐含的社会契约之下，用以界定谁创造价值、谁获得回报，以及权力结构为何合理。AI 通过放大个体杠杆、降低协调成本、提升知识的可复制性，重塑了组织内部的生产函数。然而，薪酬、晋升与决策机制，往往仍锚定在以人数、层级或资历为核心的旧模型之上。当组织内部的价值创造与价值获取发生背离，合法性便会流失，政治博弈加剧，高绩效者选择 disengage 或离开。

历史反复表明，这类错配无法仅靠叙事修复。社会契约最终总是通过结构性改变被重写。在以往的转型中，这意味着重新定义何为“贡献”，改变剩余的分配方式，并构建与物质现实相一致的新合法性叙事。在 AI 时代，这很可能意味着收入与就业的部分脱钩，从以劳动为基础的税制转向以资本与租金为核心的体系，以及将 AI 系统视为一种社会基础设施，而非纯粹的私有资产。

尤为重要的是，社会契约几乎从不被主动设计。更多时候，它是在不稳定与冲突的压力之下被重新谈判的。历史的教训不在于动荡是否能够避免，而在于代价取决于时机。那些更早适应的社会与组织，能够获得韧性；而拖延者，往往只能在极为不利的条件下被动改革。

人工智能改变的，不只是工作的方式，而是工作作为经济生活基础这一概念本身。定义现代社会的劳动型社会契约，已经承受着前所未有的压力。它的继任者，将通过理性设计诞生，还是在危机中被迫成形，仍然是一个悬而未决、却极其紧迫的问题。