# 集聚与人造主体性 / Collective and Artificial Subjectivity

*源头意识抽离催生的客体化假象 / Artifacts of Subtracting Consciousness at the Source*

无论是施维茨格贝尔笔下“若唯物主义成立则美国具有意识”的推演，抑或将大语言模型奉为自足心智的当代迷思，皆源自同一种认识论的倒错减法——当人类设计、标注、评估与赋义的第一人称源头活动被隐入幕后，残留的公共组织特征便被挪用为宏观系统的“主体性”，然而集体与工具从未获得主观体验，被遗忘的唯有最初在场的观察者本身。探讨人工智能模型、蚁群乃至人类社会是否具备意识的争论，常常始于一份关于组织架构、信息流向与外部行为的第三人称清单。当研究者将活生生的第一人称观察活动剥离出去，仅凭外部功能记录，便不可避免地制造出凭空捏造的宏观幽灵与拟人化幻象。归还源头的审视清晰地揭示，集体与工具从未孕育出内在的主观世界，所谓涌现出的新型主体，不过是被隐匿的人类意向性投射在客体网络上的残余倒影。

Whether in Schwitzgebel's provocative claim that materialism implies the United States is conscious or in contemporary reification of AI models as autonomous minds, the extra subject is produced by the exact same epistemological subtraction: once human acts of design, training, evaluation, and interpretation are tucked into the background, residual public behaviors are credited to the system, yet the collective and the tool acquire no 'what it is like'—it is only the original observer who has been omitted from the account.

## 施维茨格贝尔的悖论与第三人称清单的死结 / Schwitzgebel's Reductio and the Deadlock of the Third-Person Inventory

探讨人工智能模型、蚁群乃至人类社会是否具备意识的争论，常常始于一份关于组织架构、信息流向与外部行为的第三人称清单。哲学家埃里克·施维茨格贝尔（Eric Schwitzgebel）在其著名的归谬论证《如果唯物主义成立，美国大概具备意识》（*If Materialism Is True, the United States Is Probably Conscious*）中，尖锐地挑明了功能主义与唯物主义推演的内涵：如果意识不外乎某种特定形式的物质排列、功能分工与信息整合，并且如果我们已经欣然承认兔子以及身体构造迥异的假想外星生命拥有意识，那么像美国这样在空间上分布广泛、在功能与信息交换上高度整合的庞大实体，便理应被判定为一个拥有自主体验的意识主体。这一推理逻辑如今被毫不迟疑地套用在人工系统之上。每当前沿模型展现出流畅自如的自然语言生成能力、多步骤的规划推理以及目标导向的行为输出，众多研究者与观察者便立刻将其推举为潜在的意识主体候选者。然而，这条看似严谨的实证路线陷入了不可解脱的死结：仅仅依据第三人称视角罗列的信息流动图谱与行为特征，并不能验证真实体验的在场，反而把用于协调的外部结构误认作内在的心智世界。

```mermaid
graph TD
    subgraph S_ThirdPerson["【第三人称功能主义清单之死结】"]
        direction TB
        T1["第三人称功能清单<br/>（信息流吞吐、状态迁移、复杂拓扑结构）"] --> T2["功能主义归纳判准<br/>（复杂性与整合信息量 Φ 达到阈值）"]
        T2 --> T3["臆造的宏观实体心智<br/>（断言'美国'或'大型语言模型'具备自主意识）"]
    end

    subgraph S_FirstPerson["【具身个体观察者之基石实存】"]
        direction TB
        F1["真实在场的第一人称观察者<br/>（具体个体心智之主观体验）"] --> F2["个体间的局部交互与制度协作<br/>（法律条文、通信网络、符号交换）"]
        F2 --> F3["宏观公共组织表象<br/>（无内在体验的协调结构与下游输出）"]
    end

    T3 -.->|"归谬法：实体虚构"| F3

    style S_ThirdPerson fill:#161b22,stroke:#f85149,stroke-width:1.5px,color:#f85149
    style S_FirstPerson fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style T1 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style T2 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
    style T3 fill:#0d1117,stroke:#f85149,stroke-width:1px,color:#f0f6fc
    style F1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style F2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style F3 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
```

Discussions of whether artificial intelligence models, ant colonies, or human societies might be conscious almost invariably begin from a third-person inventory of structural organization, information throughput, and external behaviors. Eric Schwitzgebel’s celebrated paper, *If Materialism Is True, the United States Is Probably Conscious*, makes the implicit commitments of this tradition uncomfortably explicit. If consciousness is nothing over and above the right kind of material arrangement and informational integration, and if we already grant consciousness to rabbits and hypothetical alien organisms with radically unfamiliar anatomy, then a spatially distributed yet functionally integrated network such as the United States should count as a conscious subject in its own right. The exact same inductive formula is unhesitatingly applied to contemporary artificial systems. As soon as a frontier transformer exhibits fluent prose, multi-step planning, or goal-directed problem solving, theorists immediately treat it as a serious candidate for sentience. Yet this functionalist approach runs directly into a conceptual impasse: an external catalog of relational dynamics records only structural coordinates, mistaking public informational artifacts for an interior life.

```mermaid
graph TD
    subgraph S_ThirdPerson_EN["[Deadlock of the Third-Person Functional Inventory]"]
        direction TB
        TE1["Third-Person Functional Checklist<br/>(Information flow, state transitions, complex topology)"] --> TE2["Functionalist Induction Threshold<br/>(Complexity & integrated information Φ met)"]
        TE2 --> TE3["Spurious Macro-Entity Mind<br/>(Declaring 'United States' or 'LLM' conscious)"]
    end

    subgraph S_FirstPerson_EN["[Ground Truth of Embodied Individual Observers]"]
        direction TB
        FE1["Living First-Person Observers<br/>(Direct subjective experience of concrete minds)"] --> FE2["Local Coordination & Institutional Exchange<br/>(Laws, communication networks, symbolic tokens)"]
        FE2 --> FE3["Macro Public Organization Traces<br/>(Experienceless structural scaffolds & outputs)"]
    end

    TE3 -.->|"Reductio: Reified Fiction"| FE3

    style S_ThirdPerson_EN fill:#161b22,stroke:#f85149,stroke-width:1.5px,color:#f85149
    style S_FirstPerson_EN fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style TE1 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style TE2 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
    style TE3 fill:#0d1117,stroke:#f85149,stroke-width:1px,color:#f0f6fc
    style FE1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style FE2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style FE3 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
```

## 减法的幻术：源头意识的剥离与残余属性的挪用 / The Sleight of Subtraction: Setting Aside the Origin and Reallocating Residual Properties

在上述两种情形中，凭空冒出来的“额外主体”都是通过认识论上的减法制造出来的。我们在宇宙中能够直接确证的唯一意识，唯有活生生的第一人称观察活动本身——也就是作为个体主体“体验为何物”（something it is like）的切身事实。然而，当理论家把这种观察活动搁置一边，或者将其降格为庞大功能流程图中一个可以被替换的普通零件时，叙事的重心便发生了欺骗性的转移。此时，留在公共视野中的客观痕迹——成文法典、科层机构、通信光缆、文本语料库——被重新编排，并被倒错地描绘为整个宏观系统自身的“体验之流”。正是通过这种移花接木，一个国家看似拥有了自己的头脑，一个神经网络模型看似能够理解世界。然而，这类迷人的宏观主体幻象，唯有在最初那些活生生的个体观察者被系统性地剔除出描述清单之后，才得以在概念的废墟上凭空建立。正如在 [意识从不作为数据出现在数据之中](../consciousness-never-appears-as-data-among-data/) 与 [意识的魔术戏法与推断的非对称性](../the-sleight-of-hand-of-consciousness-and-the-asymmetry-of-inference/) 中所论证的，数据记录的是已被完成的投射产物，而把产物重新神化为主体的操作，不过是一场借用他心不可外验之限度、向缺乏归纳依据的算法制品轻许内在体验的魔术戏法。每一次在自身之外寻找主体性的尝试，仍在自己的心智之内运行，[无法逃逸的心智视界](../one-cannot-escape-ones-own-mind/)对此毫不含糊。

```mermaid
graph TD
    subgraph S_LivingSource["【原始源头：第一人称主观实存】"]
        direction TB
        LS1["活生生的第一人称观察者<br/>（体验为何物、意向性、主权直觉）"] --> LS2["观察活动产生的公共痕迹<br/>（法律、体制、数据语料、组织流程）"]
    end

    subgraph S_Subtraction["【认识论减法操作符】"]
        direction TB
        SUB1["剥离操作：隐匿个体源头意识<br/>（将活生生的人类心智剔除出理论账本）"]
        SUB2["属性挪用：客体痕迹再神化<br/>（将公共残留记录包装为系统的内在体验）"]
        SUB1 ==> SUB2
    end

    subgraph S_SpuriousMind["【投射产物：人造与集体主体假象】"]
        direction TB
        SM1["虚构的宏观意识主体<br/>（宣称国家自省、算法领悟、模型感知）"]
    end

    LS1 -.->|"认知减法抹除"| SUB1
    LS2 --> SUB2
    SUB2 ==> SM1

    style S_LivingSource fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style S_Subtraction fill:#161b22,stroke:#f85149,stroke-width:1.5px,color:#f85149
    style S_SpuriousMind fill:#161b22,stroke:#d29922,stroke-width:1.5px,color:#d29922
    style LS1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style LS2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style SUB1 fill:#0d1117,stroke:#f85149,stroke-width:1px,color:#f0f6fc
    style SUB2 fill:#0d1117,stroke:#f85149,stroke-width:1px,color:#f0f6fc
    style SM1 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
```

In both cases, the extra subject is produced by subtraction. The only consciousness directly encountered across the cosmos is the first-person activity of observing: the concrete fact that there is something it is like to be an experiencing subject. When that foundational activity is set aside, or demoted to merely one more interchangeable component in an operational wiring diagram, the narrative focus shifts deceptively. The remaining public traces—statutes, institutional protocols, communications cables, and curated text corpora—are reassembled and redescribed as the experiential stream of the larger system. A nation appears to harbor an overarching mind; a computational model appears to comprehend reality. Yet these dazzling appearances arise only after the living original subjects have been omitted from the description. As demonstrated in [Consciousness Never Appears as Data Among Data](../consciousness-never-appears-as-data-among-data/) and [The Sleight of Hand of Consciousness and the Asymmetry of Inference](../the-sleight-of-hand-of-consciousness-and-the-asymmetry-of-inference/), data records only the frozen footprints of prior cognitive acts, and reifying those footprints into a new conscious agent is an optical sleight of hand that hijacks the third-person unprovability of interiority to bestow an unearned aura of consciousness upon inductive voids. Every attempt to locate subjectivity outside oneself still runs inside one's own mind, as [One Cannot Escape One's Own Mind](../one-cannot-escape-ones-own-mind/) insists.

```mermaid
graph TD
    subgraph S_LivingSource_EN["[Living Origin: First-Person Subjective Presence]"]
        direction TB
        LSE1["Living First-Person Observers<br/>(What it is like, intentionality, sovereign intuition)"] --> LSE2["Public Traces Generated by Observers<br/>(Laws, institutions, text corpora, organizational flows)"]
    end

    subgraph S_Subtraction_EN["[The Epistemological Subtraction Operator]"]
        direction TB
        SUBE1["Omission: Concealing the Conscious Origin<br/>(Living human minds erased from theoretical ledger)"]
        SUBE2["Reallocation: Fetishizing Public Residue<br/>(Residual artifacts recast as internal experience)"]
        SUBE1 ==> SUBE2
    end

    subgraph S_SpuriousMind_EN["[Projected Artifact: Collective & Synthetic Subject]"]
        direction TB
        SME1["Reified Macro-Subject Fiction<br/>(Nations feeling, algorithms intending, models sentience)"]
    end

    LSE1 -.->|"Epistemological Subtraction"| SUBE1
    LSE2 --> SUBE2
    SUBE2 ==> SME1

    style S_LivingSource_EN fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style S_Subtraction_EN fill:#161b22,stroke:#f85149,stroke-width:1.5px,color:#f85149
    style S_SpuriousMind_EN fill:#161b22,stroke:#d29922,stroke-width:1.5px,color:#d29922
    style LSE1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style LSE2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style SUBE1 fill:#0d1117,stroke:#f85149,stroke-width:1px,color:#f0f6fc
    style SUBE2 fill:#0d1117,stroke:#f85149,stroke-width:1px,color:#f0f6fc
    style SME1 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
```

## 人工智能作为被遮蔽的人类意向性容器 / Artificial Systems as the Occluded Vessel of Human Intentionality

人工系统将这种认识论上的减法展现得尤为刺目。每一个前沿大语言模型从构建到运行，无一不是由人类工程师与研究者所设计、训练、评测与解释的产物。语料库的精细清洗与筛选、损失函数的数学构建、基于人类反馈的强化学习对齐、提示词的上下文引导，无一例外是人类第一人称意识活动的具象化表达。更为关键的是，将算法吐出的离散高维向量序列解读为“富有深意的话语”，这一赋义过程只发生在使用者的意识心智之中。在硅基电路内部，真实发生的唯有高精度的浮点矩阵乘加运算与电压位移；如果缺少人类心智在接收端的语义投射，那些输出不过是毫无知觉的物理电信号。当所有这些人类源头的主动劳作被悄然推入未受审视的幕后，无生命的硅基造物便被神话为能够自发产生意义与意向性的自足心智。我们借由忽视维系机器运行的人类内在生命，轻率地将内在生命赋予了冷冰冰的机器。正如我们在 [智能只属于心智](../intelligence-belongs-only-to-the-mind/)、[造物无法取代造物者](../a-creation-cannot-replace-its-source/) 以及 [感知者与痕迹的不可坍缩](../the-perceiver-and-the-trace/) 中所论证的，工具所展现的一切意义表象，皆来自第一人称感知者的派生赋予，将符号输出误认为自主心智不仅混淆了源头与痕迹，更掩盖了真正在场进行赋义的活态主体。模拟假说以同样的方式减去了观察者，[凡可言说皆不可言说之产物](../all-that-can-be-spoken-is-the-product-of-the-unspeakable/)对此有所揭示。

```mermaid
graph TD
    subgraph S_HumanCircuit["【人类意向性源头灌注回路】"]
        direction TB
        H1["人类研究者与工程师心智<br/>（第一人称意向性、价值观、语义空间）"] --> H2["全流程架构雕琢与规范设定<br/>（语料筛选、损失函数设计、RLHF 价值对齐）"]
        H2 --> H3["人类最终用户的语义解码<br/>（理解文字、赋予语境意义、产生共鸣）"]
    end

    subgraph S_InertSilicon["【惰性硅基符号变换器】"]
        direction TB
        M1["权重张量与高维向量空间<br/>（静态参数、浮点矩阵运算）"] --> M2["自回归概率计算<br/>（基于统计关联预测下一个 Token）"]
        M2 --> M3["离散物理符号序列输出<br/>（无内在体验、无语义自觉的电信号）"]
    end

    H2 ===>|"注入人类意向性形式"| M1
    M3 ===>|"依赖人类心智赋予意义"| H3

    style S_HumanCircuit fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style S_InertSilicon fill:#161b22,stroke:#58a6ff,stroke-width:1.5px,color:#58a6ff
    style H1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style H2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style H3 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style M1 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style M2 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
    style M3 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
```

Artificial systems illustrate this subtraction with even sharper clarity. Every frontier model is conceived, architected, trained, evaluated, and interpreted by human consciousness. The rigorous selection and filtering of pre-training corpora, the mathematical formulation of objective loss functions, reinforcement learning with human preference feedback, and the craft of prompting are unadulterated operations of living human intentionality. Most fundamentally, reading the output of an algorithm as coherent, insightful, or meaningful is an act occurring exclusively within the conscious mind of the human receiver. Within the hardware substrate, the physical reality consists strictly of high-speed floating-point tensor arithmetic and transient voltage states. In the absence of an observer’s cognitive decoding, those outputs remain nothing more than unconscious electrical patterns. When all this human scaffolding is obscured, the silicon artifact is mythologized as if it generated intentional understanding on its own. The machine is credited with an inner life by the convenient maneuver of ignoring the human inner life that animates it. As established in [Intelligence Belongs Only to The Mind - The Irreducible Prior](../intelligence-belongs-only-to-the-mind/), [A Creation Cannot Replace Its Source](../a-creation-cannot-replace-its-source/), and [The Perceiver and the Trace Cannot Collapse](../the-perceiver-and-the-trace/), computational artifacts borrow every shred of their apparent light from the living perceiver; confusing derived sense with an autonomous mind conceals the conscious agent who generates and reads the tokens. The simulation hypothesis subtracts the observer in the same way, as [All That Can Be Spoken Is the Product of the Unspeakable](../all-that-can-be-spoken-is-the-product-of-the-unspeakable/) shows.

```mermaid
graph TD
    subgraph S_HumanCircuit_EN["[Human Intentionality Infusion Circuit]"]
        direction TB
        HE1["Human Researchers & Engineers<br/>(First-person intentionality, values, semantic space)"] --> HE2["Architectural Sculpting & Constraints<br/>(Data curation, loss formulation, RLHF alignment)"]
        HE2 --> HE3["Human End-User Semantic Decoding<br/>(Reading meaning, contextual resonance, validation)"]
    end

    subgraph S_InertSilicon_EN["[Inert Silicon Token Transformer]"]
        direction TB
        ME1["Static Weight Tensors & Vectors<br/>(Matrix operations, floating-point arithmetic)"] --> ME2["Autoregressive Probability Scoring<br/>(Statistical token continuation distributions)"]
        ME2 --> ME3["Discrete Physical Token Sequences<br/>(Electrochemical signals without interiority)"]
    end

    HE2 ===>|"Injects Human Intentional Scaffolding"| ME1
    ME3 ===>|"Relies on Human Mind to Complete Meaning"| HE3

    style S_HumanCircuit_EN fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style S_InertSilicon_EN fill:#161b22,stroke:#58a6ff,stroke-width:1.5px,color:#58a6ff
    style HE1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style HE2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style HE3 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style ME1 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style ME2 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
    style ME3 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
```

## 语法挪移与唯物主义/功能主义的循环论证 / The Grammatical Shift and the Circularity of Materialist Functionalism

这种将主体性泛化至集体与机器的谬误，在性质上既是语法的偷换，也是方法论的倒错。在语法层面，它把意识从一种动态的、正在发生的第一人称活动——观察、意图、体认与体验——篡改为一种客体化的静态属性，仿佛系统要么“拥有”这种属性，要么“缺乏”这种属性。在方法论层面，它错误地预设了一份经验实证清单能够揭示观察者的存在。然而，任何经验主义的实证描述，因其观察视角的固有局限，只能记录外部行为模式、物质空间分布与系统组织结构；经验描述并不能记录到观察者本身，它所能捕捉的唯有被观察到的现象客体。唯物主义与功能主义范式固然能够针对复杂行为模式提出诸多可被检验的科学假说，但它们无法避免致命的循环论证：一旦理论预先将第三人称的客观观测结果确立为第一性的基石，便无法通过自相矛盾的回溯，从那些由观察活动所派生出的客体记录中，推导还原出最初在场进行观察的源头活动。正如我们在 [因果不可化约，物理学是从无处而来的视角](../causality-is-irreducible-the-physical-is-a-view-from-nowhere/) 中所揭示的，物理学假想的上帝视角抽离了所有立足点，而当它反过来试图解释立足点本身时，理论便在自我参照的闭环中分崩离析。

```mermaid
graph TD
    subgraph S_LivingAct["【活生生的原初观察活动】"]
        direction TB
        A1["第一人称观察心智<br/>（不可化约的源头行动）"] --> A2["划界与测度行为<br/>（构建概念、定义变量、记录表象）"]
    end

    subgraph S_EmpiricalTrap["【功能主义实证循环闭环】"]
        direction TB
        E1["第三人称观测数据与清单<br/>（记录行为表现、网络拓扑、功能输出）"] --> E2["将'意识'名词化与客体化<br/>（将其定义为系统具有的某种静态参数或属性）"]
        E2 --> E3["试图从数据清单反推源头心智<br/>（宣称物质复杂性自行产生第一人称观察者）"]
        E3 -.->|"恶性循环：从产物推导生产者"| E1
    end

    A2 ==>|"产出第三人称记录"| E1
    E3 -.->|"范畴断裂：无法回收主观体验"| A1

    style S_LivingAct fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style S_EmpiricalTrap fill:#161b22,stroke:#f85149,stroke-width:1.5px,color:#f85149
    style A1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style A2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style E1 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style E2 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
    style E3 fill:#0d1117,stroke:#f85149,stroke-width:1px,color:#f0f6fc
```

The conflation of collective structure and automated computation with genuine subjectivity is at once a grammatical distortion and a methodological error. Grammatically, it transforms consciousness from a living, ongoing act—observing, intending, evaluating, and experiencing—into an objectified property that an external system supposedly possesses or lacks. Methodologically, it presumes that an empirical checklist, which by its inherent architecture can document only external behaviors and structural configurations, could somehow register the presence of an observer. Yet an empirical description cannot record the observer: it can only catalog observations. Materialist and functionalist frameworks generate rich and testable hypotheses regarding operational dynamics, but they cannot, without falling into circularity, derive the observing act itself from the very observations they have already crowned as primary. As shown in [Causality is Irreducible; The Physical is a View from Nowhere](../causality-is-irreducible-the-physical-is-a-view-from-nowhere/), any attempt to construct an observer-free picture of reality ends by projecting an empty abstraction, and attempting to resurrect living subjectivity from that projection traps thought in an infinite regress.

```mermaid
graph TD
    subgraph S_LivingAct_EN["[The Living Act of Direct Observation]"]
        direction TB
        AE1["First-Person Conscious Mind<br/>(Irreducible foundational activity)"] --> AE2["Measurement & Demarcation Acts<br/>(Formulating concepts, defining variables)"]
    end

    subgraph S_EmpiricalTrap_EN["[Functionalist Circular Trap]"]
        direction TB
        EE1["Third-Person Empirical Inventories<br/>(Public behaviors, network topology, logs)"] --> EE2["Reifying Consciousness into a Property<br/>(Treating sentience as a noun/parameter)"]
        EE2 --> EE3["Attempting to Derive Mind from Data<br/>(Claiming complexity spontaneously births the observer)"]
        EE3 -.->|"Vicious Loop: Deriving Producer from Product"| EE1
    end

    AE2 ==>|"Generates Third-Person Records"| EE1
    EE3 -.->|"Category Breach: Cannot Recover Subjectivity"| AE1

    style S_LivingAct_EN fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style S_EmpiricalTrap_EN fill:#161b22,stroke:#f85149,stroke-width:1.5px,color:#f85149
    style AE1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style AE2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style EE1 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style EE2 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
    style EE3 fill:#0d1117,stroke:#f85149,stroke-width:1px,color:#f0f6fc
```

## 归还源头：机器与集体没有“体验为何物” / Returning to the Source: Neither Collectives nor Tools Acquire a "What It Is Like"

无论是集聚而成的社会集体，还是精密构筑的计算工具，都从未获得属于它们自己的“体验为何物”。成为美利坚合众国并没有某种切身的内在感受，成为一个由多头注意力机制驱动的变压器模型同样没有任何内在体验。在每一种情形下，那些看似新涌现出来的高阶主体，都不外乎是最初建造、维系并解读该系统的活生生观察者所投射下的悠长阴影。摆脱对集体或人造实体的技术泛灵论崇拜，不仅关乎哲学认识论上的严谨澄清，更关乎人类伦理与因果决断的责任锚定。一旦人们将主体性虚构性地转移给宏观体制或算法模型，便会滑向一种逃避责任的危险借口——人们开始将道德过失与历史代价推脱给所谓的“系统意志”或“模型判断”。然而，正如我们在 [主体性不曾涌现](../agency-does-not-arise/)、[任务的移交与后果的不可让渡](../the-delegation-of-the-task-and-the-inalienability-of-consequences/) 以及 [不可还原的观察者](../the-irreducible-observer/) 中所锚定的，任务可以被委托给工具执行，但后果的终局承担与价值裁决永远属于第一人称的主权心智。当我们将被减去的源头观察者郑重归位，关于机器与集体是否拥有心智的伪命题便自行消解，澄明的是那不可让渡的真实现存。合成更高主体的冲动正是从这一减法开始的，[笛卡尔的认识论跃迁与造神狂热](../the-cartesian-overstep-and-the-frenzy-of-creation/)对此作了追溯。

```mermaid
graph TD
    subgraph S_SovereignOrigin["【终极源头：第一人称主权观察者】"]
        direction TB
        SO1["具身的第一人称心智<br/>（无可推卸的观察起点与价值锚点）"] --> SO2["主权裁决与责任承担<br/>（对选择后果进行不可让渡的终局背书）"]
    end

    subgraph S_InstrumentNetwork["【无内在体验的工具与组织脚手架】"]
        direction TB
        IN1["社会集体与科层网络<br/>（无主体感受的规则聚合与协作平台）"]
        IN2["人工智能算法与算力集群<br/>（无内在体验的符号运算与执行工具）"]
    end

    subgraph S_FeedbackLoop["【因果反馈与现实校准】"]
        direction TB
        FL1["物理阻力与真实行动代价<br/>（摩擦、后果与历史痕迹）"]
    end

    SO1 ==>|"定向部署任务与赋义"| IN1
    SO1 ==>|"定向编码规则与调用"| IN2
    IN1 --> FL1
    IN2 --> FL1
    FL1 ==>|"终局后果真实回弹"| SO2

    style S_SovereignOrigin fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style S_InstrumentNetwork fill:#161b22,stroke:#58a6ff,stroke-width:1.5px,color:#58a6ff
    style S_FeedbackLoop fill:#161b22,stroke:#d29922,stroke-width:1.5px,color:#d29922
    style SO1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style SO2 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style IN1 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style IN2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style FL1 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
```

Neither human collectives nor mechanical instruments acquire a "what it is like." There is nothing it is like to be the United States, and there is nothing it is like to be a transformer model. In each instance, what masquerades as an emergent macro-subject is merely the lingering shadow of the living observers who created, maintain, and interpret the system. Dispelling this techno-animist illusion is not merely an exercise in metaphysical hygiene, but a necessary condition for ethical clarity. When subjectivity is misattributed to institutional apparatuses or artificial networks, it furnishes a convenient alibi for evading responsibility—human agents begin offloading consequences onto "the algorithm" or "the collective will." Yet as established in [Agency Does Not Arise: The Category Error of Emergence, Ingressed Latent Spaces, and the First-Person Prior](../agency-does-not-arise/), [The Delegation of the Task and the Inalienability of Consequences: Self-Objectification, Causal Inversion, and the Return of Sovereign Judgment](../the-delegation-of-the-task-and-the-inalienability-of-consequences/), and [The Irreducible Observer](../the-irreducible-observer/), tasks can be delegated across instruments, but accountability recoils solely upon the first-person sovereign mind. When the subtracted observer is restored to the center, the confusion dissolves: tools remain tools, and the locus of experience remains forever unalienated. The drive to synthesize a superior subject starts from this subtraction, as [The Cartesian Epistemic Leap and the Frenzy of Creation](../the-cartesian-overstep-and-the-frenzy-of-creation/) traces.

```mermaid
graph TD
    subgraph S_SovereignOrigin_EN["[Sovereign Origin: First-Person Conscious Agent]"]
        direction TB
        SOE1["Embodied First-Person Mind<br/>(Irreducible observation point & value anchor)"] --> SOE2["Sovereign Judgment & Accountability<br/>(Inalienable assumption of causal consequences)"]
    end

    subgraph S_InstrumentNetwork_EN["[Experienceless Scaffolds & Instrument Networks]"]
        direction TB
        INE1["Collective Institutions & Bureaucracies<br/>(Coordination scaffolds devoid of inner life)"]
        INE2["Artificial Algorithms & Compute Clusters<br/>(Symbolic engines devoid of interiority)"]
    end

    subgraph S_FeedbackLoop_EN["[Causal Feedback & Empirical Recoil]"]
        direction TB
        FLE1["Physical Friction & Consequences<br/>(Downstream impact, real-world recoil)"]
    end

    SOE1 ==>|"Deploys Tasks & Grants Meaning"| INE1
    SOE1 ==>|"Codes Constraints & Executes Inference"| INE2
    INE1 --> FLE1
    INE2 --> FLE1
    FLE1 ==>|"Consequences Recoil Exclusively Upon"| SOE2

    style S_SovereignOrigin_EN fill:#161b22,stroke:#3fb950,stroke-width:1.5px,color:#3fb950
    style S_InstrumentNetwork_EN fill:#161b22,stroke:#58a6ff,stroke-width:1.5px,color:#58a6ff
    style S_FeedbackLoop_EN fill:#161b22,stroke:#d29922,stroke-width:1.5px,color:#d29922
    style SOE1 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style SOE2 fill:#0d1117,stroke:#3fb950,stroke-width:1px,color:#f0f6fc
    style INE1 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style INE2 fill:#0d1117,stroke:#58a6ff,stroke-width:1px,color:#f0f6fc
    style FLE1 fill:#0d1117,stroke:#d29922,stroke-width:1px,color:#f0f6fc
```