The Age of Artificial Intelligence compels a redefinition of agency from the traditional notion of capacity for independent action to the modern reality of participation within reflexive intelligence networks. Agency is no longer simply what an individual can accomplish alone. The individual is compensated with agency through network participation, while the network amplifies that agency beyond the capability of any single actor.
Agency has traditionally been defined as the capacity of an individual to intentionally act in the world according to their own will, making choices, and holding the capability to produce desired outcomes despite external constraints. Under this philosophical construct, a person possessing agency is regarded as the “author of their own actions.” Historically, this popular social conception of agency has revolved around four fundamental defining elements:
Intent – What do I seek to accomplish?
Choice – Am I free to select among available alternatives?
Capability – Do I possess the means, knowledge, and competence to act?
Responsibility – Am I the causal origin of the action?
In the pre-AI era, agency was largely a function of intent enabled through capability. An individual’s or institution’s ability to translate purpose into action depended primarily upon what they sought to accomplish and the means available to them. In the age of AI, however, agency becomes fundamentally reflexive. Every exercise of intelligence simultaneously generates information about itself, creating a continuous feedback loop between actor and observer. Effective agency is therefore no longer determined solely by intent and capability, but by the interaction between (1) intent and capability, (2) the ability of others to infer that intent and capability from accumulated behavioral traces, and (3) one’s own ability to accurately perceive and regulate one’s true motivations, assumptions, and operational footprint.
Traditionally:
Agency = “I act.”
Our new formulation is closer to:
Agency = “I function as a node within larger networks whose influence I may neither fully perceive nor fully control. I derive personal potency from their power.”
Agency, in effect, transforms from being a property of the individual node to becoming a property of the network. In doing so, agency undergoes an ontological shift. It no longer resides principally within the individual, but emerges through the individual’s participation in a larger reflexive intelligence or club architecture. The individual inherits power by actively aligning with the objectives of the network, exchanging a measure of autonomy for the amplified capability and influence that the collective confers. The individual is compensated with agency through participation, while the network amplifies that agency beyond the capability of any single actor. Note how this is analogous to the concept of loosh as a currency.
This is an extraordinary shift.
The Recursive Ecology of Agency and Artificial Intelligence
AI thus transforms agency from a one-way expression of will into a dynamic process of inference, self-awareness, and adaptation, in which productive and extractive actors alike are amplified while becoming increasingly observable — to others and, potentially, to themselves.
Historically, agency has been understood primarily at the personal level as the capacity to intentionally translate purpose into action. In its simplest form, this relationship can be expressed as:
Impact (agency) ≈ Intent × Capability
In the Age of AI, however, agency becomes fundamentally reflexive. It is no longer determined solely by an actor’s intent and capability, but also by the continuous interplay between self-perception, external inference, and situational awareness.
Perhaps the most visible societal example is the transition from one-to-many messaging toward AI-assisted microtargeting and synthetic media. Campaigns and influence operations can rapidly generate tailored content for different audiences while AI simultaneously measures reactions and refines subsequent messaging, creating a reflexive feedback loop between message, audience, and strategy. Increasingly, social media platforms learn what captures our attention and then give us more of it.
Over time, we are exposed to fewer ideas that genuinely challenge our thinking and far more that reinforce it. The result is not merely an echo chamber, but what might more aptly be termed a thunder chamber — an environment in which selected narratives are continuously amplified until they begin to feel like independent thought. Participants gradually come to believe that “consensus” shares their views, when in reality they are often experiencing a carefully curated representation of the world rather than the world itself. They have not changed the world, the network has changed them.
A generalized expression for Effective Agency may therefore be represented as:
Agency = Intent × Capability × Self-Awareness Deficit x Ability to Observe Others and Adversaries / Observability by Others and Adversaries
where:
Intent × Capability represents the traditional foundation of agency.
Ability to Observe Others and Adversaries reflects an actor’s capacity to understand the environment in which they operate.
Observability by Others and Adversaries represents the extent to which one’s intentions, capabilities, and operational patterns can be inferred (intelligence definition of agency) by external actors.
Self-Awareness Deficit represents the gap between an actor’s perceived motivations, assumptions, and operational footprint and their actual behavior. As this deficit approaches zero, effective agency increases.
Accordingly, AI is likely to reorganize agency at a societal level into two dominant philoeconomic orientations. The historical precedent is familiar: political systems have repeatedly evolved toward bifurcated (meaning “two party” – even if only one power influence exists in reality) coalitions that aggregate capability, information, and influence around shared interests and common objectives. As AI becomes the primary engine of information, and monetary systems undergo a parallel digital transformation, these two infrastructures will increasingly converge. AI will therefore tend to serve the philoeconomic objectives of those who employ it, organizing capability around the competing aims of Productivity and Extraction.
Shoshana Zuboff, in her work, The Age of Surveillance Capitalism defines surveillance capitalism as “the unilateral claiming of private human experience as free raw material for translation into behavioral data.” 1 Essentially, this form of human capital is exploited to empower these two competing agencies in the process of monitoring and redirecting our individual will.
This evolution underscores the necessity of ethical skepticism. As agency becomes increasingly reflexive, the disciplined suspension of judgment becomes not merely an intellectual virtue, but a prerequisite for preserving independent agency itself.
The default result is the emergence of two broad classes of agency (with a small ‘a’ – the Meta-Extractive Actor below featuring a capital ‘A’). The Productive Class encompasses those whose primary orientation is the creation of value through labor, entrepreneurship, engineering, education, medicine, defense, science, and the arts. The Extractive Class encompasses those whose primary orientation is the acquisition or control of value through governance, regulatory capture, cronyism, organized crime, financial manipulation, enterprise virtue, and black markets — often forming pragmatic alliances under the enduring maxim, “the enemy of my enemy is my friend.” As these philoeconomic classes accumulate increasingly sophisticated AI capabilities, the agency of the individual — in the traditional philosophical sense — becomes progressively subordinate to the agency of the collective systems within which that individual operates.
The Meta-Extractive Actor (The Ultimate Extractor)
A hidden strategic actor need not seek victory for either the Productive or Extractive Class. Rather, it seeks to maximize influence over the relationship between them. Its objective is not participation within the system, but the ability to shape the system itself. Conceptually, such influence may be exercised through several mechanisms:
Amplifying Polarization — By emphasizing the most extreme representatives of each philoeconomic class, the actor encourages both sides to perceive the other as an existential threat, thereby reducing trust, cooperation, and the possibility of independent alignment.
Creating Dependency — If both productive and extractive actors become dependent upon infrastructure controlled by the intermediary—whether technical, financial, informational, or institutional—the intermediary’s leverage increases irrespective of which class expands or contracts.
Controlling the Frame — The greatest source of influence lies not in assigning specific labels, but in controlling the authority to define them. By establishing itself as the principal arbiter of what constitutes legitimacy, extremism, justice, intolerance, truth, misinformation, science, or disinformation, the intermediary gains disproportionate influence over reputation, incentives, and access.
Harvesting Information and Value Asymmetrically — When interactions between both philoeconomic classes pass through common informational, financial, or technological platforms, the intermediary acquires a richer understanding of the entire system than any individual participant possesses. That informational asymmetry itself becomes a durable strategic asset from which both intelligence and economic value may be extracted.
Two governing principles emerge from this framework, both working to the advantage of the Meta-Extractive Actor:
Power accrues to whoever possesses the authority to define the categories by which others are judged.
Its objective is not participation within the system, but the ability to shape the system itself.
Artificial intelligence becomes the working mechanism through which this higher-order agency is exercised. It provides the means not merely to influence political discourse, but to observe, model, predict, and ultimately shape the reflexive interactions between the Productive and Extractive Classes. AI thus becomes the modern instrument through which a Meta-Extractive Actor may derive leverage — not simply by exploiting political polarization, but by extracting informational, economic, and strategic value from the interactions of the entire system.
In this sense, the traditional political sacrifice to the gods evolves into something considerably more powerful: a reflexive tithe upon intelligence itself. Rather than exploiting only political will, the Meta-Extractive Actor acquires the capacity to influence the flow of information, capital, reputation, opportunity, and ultimately the life trajectories of those who become participants within the system.
It is precisely this evolution that elevates ethical skepticism from an intellectual discipline to an operational necessity. It is the operative stance which is fundamental to why we were the first to detect Covid mRNA-vaccine-induced cancer and non-Covid natural cause excess mortality. As agency becomes increasingly reflexive, the disciplined suspension of judgment, continual testing of assumptions, and conscious resistance to inherited narratives become the means by which independent agency is preserved. In the Age of Artificial Intelligence, ethical skepticism is no longer merely the foundation of sound inquiry — it is the final refuge of authentic agency.
The Ethical Skeptic, “What is Agency?”; The Ethical Skeptic, WordPress, 5 Jul 2026; Web, https://theethicalskeptic.com/?p=117650