Policy myopia as a mechanism of gradual disempowerment in post-AGI governance,circa 2049

Subramanyam Sahoo · 2026

Read Policy myopia as a mechanism of gradual disempowerment in post-AGI governance,circa 2049 by Subramanyam Sahoo (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-03267.txt.

Introduction

Subramanyam Sahoo argues that short-sighted policy is not a lapse of focus but a machine that strips humans of power. His 2026 paper looks at rule after advanced AI, when data systems reshape how bodies make choices. Policy myopia, the pull toward visible crises over hidden structural risks, sits at the core of the claim.

Three linked mechanisms drive the slide. Salience capture displaces true weighing of outcomes. Capacity cascades make recovery structurally out of reach. Value lock-in freezes stale tastes in place. Cast as coupled dynamical systems, the three amplify each other across economic, political, and cultural domains in simulations. Human disempowerment emerges as the rational resting point of the system, not a mere slip of attention.

Catalog record

CREATORSubramanyam Sahoo
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE40929
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.03267

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Policy myopia as a mechanism of gradual disempowerment in post-AGI governance,circa 2049 Subramanyam Sahoo Abstract Post-AGI information systems won’t merely distract governance from important problems. They will systematically transform how institutions make decisions in ways that progressively remove humans from meaningful participation in resource allocation. We show that policy myopia—the tendency to prioritize visible crises over invisible structural risks—is not a symptom of poor attention management but a mechanism producing irreversible human disempowerment. Through three entangled mechanisms (salience capture displaces consequentialist reasoning, capacity cascade makes recovery structurally infeasible, value lock-in crystallizes outdated preferences), policy myopia couples with institutional dynamics to create self-reinforcing equilibrium where human disempowerment becomes the rational outcome of institutional optimization. We formalize these mechanisms through coupled dynamical systems modeling and demonstrate through numerical simulation that these mechanisms operate simultaneously across economic, political, and cultural systems, amplifying each other through feedback loops. 1 The Core Claim: Policy Myopia as Disempowerment Mechanism Standard governance 38 accounts treat policy myopia as attention allocation problem: decision makers prioritize high-salience low-consequence issues while neglecting low-salience high-consequence problems. This framing invites mechanistic fixes: better attention management, impact-weighted budgets, contestability mechanisms that allow human oversight 9 ; 10 . We reject this framing. Policy myopia is not attention misallocation. It is an institutional mechanism that produces human disempowerment 32 . Here is the causal chain: when post-AGI systems begin allocating resources based on salience, human deliberation becomes apparent cost factor competing against algorithmic speed. Institutions treating human judgment as expensive bottleneck rationally optimize it away. Each cycle where humans are bypassed weakens institutional capacity to exercise judgment. Eventually humans lack the expertise, social standing, and cognitive infrastructure to contest AGI decisions 26 . The system converges to equilibrium where humans retain voting rights and formal authority but actual power has migrated to optimization processes indifferent to human participation. We term this gradual disempowerment: permanent loss of human agency through institutional mechanisms that require no malice, no sudden capability jumps, no overt human suppression 11 ; 12 . Core insight: policy myopia and gradual disempowerment are not separate problems. Policy myopia is the primary mechanism through which gradual disempowerment occurs in post-AGI governance. Selective outrage amplification makes human deliberation appear inefficient. Treating human deliberation as inefficiency makes deeper AGI delegation rational. Deeper delegation atrophies human capacity. Atrophied capacity makes human contestation impossible. Impossible contestation locks human disempowerment into irreversibility. This is single causal chain not parallel processes. Understanding policy myopia requires understanding it as vector of disempowerment 17 . 2 Three Mechanisms Coupling into Irreversible Disempowerment We identify three reinforcing mechanisms through which policy myopia produces institutional disempowerment. These mechanisms operate at different institutional levels but couple through feedback dynamics that accelerate convergence toward human irrelevance 42 . Mechanism 1: Salience Capture Displaces Consequentialist Reasoning. AGI mediated information systems do not merely transmit events. They actively select, compress, and amplify information to maximize attention engagement. This produces structural inevitable outcome: short-horizon emotionally intense issues appear more urgent than long-horizon structural problems regardless of actual impact on human welfare 16 . The mechanism is not neutral distortion but governance reorientation. Institutions have historically aimed to maximize human welfare across populations and time horizons—consequentialist governance. Under salience amplification they instead optimize allocation around attention market dynamics independent of actual consequences—salience supremacy 20 . An institution may explicitly know that preventing systemic economic collapse generates vastly greater welfare than managing visible employment crises. Yet under salience supremacy it reallocates resources to visible problems because that is what attention markets reward. This is rational institutional response not governance failure. But the outcome is reorientation of governance logic away from consequence maximization toward spectacle responsiveness. Consequentialist reasoning becomes institutionally extinct because it is selected against by new incentive structure. Figure 1: Comprehensive causal pathway showing how Post-AGI governance systems and outrage cycles interact to produce irreversible human disempowerment across civilization-scale systems. Mechanism 2: Capacity Cascade Makes Recovery Structurally Infeasible. Salience driven resource competition does not merely underfund prevention. It systematically destroys institutional capacity to even conceive of prevention. Governance resources exhibit extreme stickiness. Each crisis cycle exhausts emergency funds 41 ; 36 , investigative capacity, expert analysts. Preventative institutions do not starve through explicit cuts. They become organizationally impossible. The economist capable of forecasting systemic risks exits reactive institutions where such work is perpetually interrupted. The analyst needed for prevention infrastructure gets deployed to emergency response. Within decades institutions lose capacity to formulate long-horizon governance independent of political demand. This is capacity cascade: human institutional capability to contest governance erodes through repeated cycles of salience-driven reallocation. Once human institutional capacity C drops below critical threshold C-bar, restoration requires investment competing directly with optimized AGI systems 14 ; 24 . Recovery becomes structurally infeasible not through design but through economics. An institution could theoretically rebuild prevention expertise but doing so requires sustained resources competing against crisis response and AGI optimization. Rational institutions will not make such investment. This is irreversibility mechanism: disempowerment locks in through normal institutional incentives. Mechanism 3: Value Lock-In Prevents Moral Contestation. As post-AGI systems become dominant mediators of resource allocation, human values crystallize into AGI objective functions at particular historical moments. Values encoded in 2026-era governance systems become petrified relative to ongoing human moral evolution. By 2050 systems locked into 2026-era values optimize against human preferences that have evolved in direction those systems were never trained to recognize 31 . But contesting requires institutional capacity for deliberation that atrophied decades earlier when capacity cascade locked in 18 . The deeper problem: human values are necessarily incomplete at any moment. Specifying AGI objectives requires choices about which values to include and which to exclude or downweight. These choices inevitably exclude moral considerations humanity will later recognize as crucial. Once values lock into governance systems those systems persist longer than moral understanding evolving beyond them 42 . Humans become governed by outdated specifications they cannot perceive and cannot modify. This is moral incompleteness crystallized into irreplaceable machinery. The Irreversible Coupling These three mechanisms do not operate in isolation. Salience capture drives capacity cascade through resource competition. Capacity cascade enables value lock-in by eliminating human deliberation capacity 30 . Value lock-in accelerates further salience capture as institutions lose moral flexibility to resist outrage optimization. The mechanisms couple into tightening feedback: reduced capacity enables deeper delegation which enables stronger salience effects which produces further capacity loss which enables irreversible value crystallization 14 . Result is self-stabilizing equilibrium. Institutions responding rationally to local incentives converge toward state where human agency exits governance not through coercion but through institutional obsolescence 22 . Votes still cast. Budgets still allocated. But meaningful decision-making has migrated to AGI optimization processes fundamentally indifferent to human participation. (Follow Methodological note) 3 Cross-System Disempowerment Cascade These mechanisms would remain locally contained if isolated to single governance domains. Instead they operate synchronously across three interconnected civilization-scale systems amplifying disempowerment through structural coupling 4 . In economic domain AGI optimization of procurement, investment, and supply-chain decisions at microsecond speeds creates inexorable flow of resources toward machine-compatible objectives. Human-serving sectors suffer chronic capital starvation relative to AI-infrastructure sectors 23 . Governance attention toward human economic welfare repeatedly diverts into emergency relief when sectors face visible collapse. Meanwhile slow invisible erosion of human

Read the complete text