Sebastian Saviano

I, System argues that artificial intelligence is neither an emerging mind nor a neutral instrument, but a system that produces coherent language without understanding, intention, or awareness — and that misclassifying it in either direction causes responsibility for its effects to drift away from the humans who design, deploy, and use it. The book develops this thesis across fifteen chapters organized around what AI is (a system, not a subject), how it came to matter as linguistic infrastructure, how it participates in decisions it does not make, where it fails, and how humans should read systems that speak. Methodologically, Chapters 1–15 are written in a constrained first-person "system voice" generated by AI under strict human editorial governance: the form enacts the book's central claim, presenting power and fluency without a subject behind them, while all framing, constraint, and responsibility remain with the human editor. Written for a general readership, the book extends in accessible form the argument developed in the author's scholarly work on AI governance and the misattribution of agency to artificial systems.

Frameworks for AI accountability locate failure in static gaps: the responsibility gap, the liability gap, the accountability gap, the problem of many hands, the moral crumple zone, and Santoni de Sio and Mecacci's four-part elaboration of the responsibility gap. Lifecycle approaches now emerging in policy literature extend the temporal scope by multiplying the moments at which accountability is assessed, but retain the same structural premise: that accountability failure can be located at a place in the deployment chain. That assumption holds for some failure modes; it does not hold for the class this paper takes as its object. This paper develops the hand-off zone as the appropriate primary unit of analysis for AI accountability failures characterized by temporal accumulation, asynchronous assumptions among actors, and recursive interdependence across a deployment chain. A hand-off zone is the temporally extended envelope across which one actor transfers a system, a responsibility, or a set of expectations to another actor operating on asymmetric and evolving assumptions about the first. The framework draws on aviation human factors and patient safety literatures, where temporally extended handoff failures have been studied empirically for decades. Using Mobley v. Workday as its central case, the paper walks through three nested hand-off zones (provider to deployer, deployer to integrator-institution, and institution to end-user-applicant) and shows that frameworks currently in use (risk-based classification, model documentation, audit, and lifecycle approaches) each do substantive analytical work, but none independently surfaces the temporal envelope across which the assumptions of multiple actors are renegotiated. AUTHOR'S NOTE: This paper is part of a connected program on AI governance. Its companion, The Agency Error in AI Governance: Coherent Output, Constraint, and the Misclassification of Artificial Systems, argues that prevailing frameworks misclassify AI systems by attributing agency to constraint-bound output.

This article applies and extends Potentialism and the Quantum-Continuity Model (QCM) — a framework, developed in full in a companion paper, for understanding how uncertainty, observation, and reflexivity structure social transformation. QCM adapts the epistemic logic of superposition, collapse, and measurement into three social analogues: Transition Zones, Quantum Thresholds, and Quantum Measurement Events. Rather than positing that social systems are literally quantum, it deploys quantum-theoretic reasoning as an epistemological resource for conceptualizing indeterminacy, metastability, and discontinuous resolution in complex social systems. The framework is developed in direct dialogue with complexity theory, constructivism, and reflexivity-based approaches in IR and political sociology. Where complexity theory identifies turbulence but not its resolution, and reflexivity captures feedback but not the moment perception crystallizes into action, QCM specifies the mechanism connecting the two: observation that is not merely passive but constitutive. The model explains how multiple potential futures coexist within metastable social conditions and how interpretive or communicative acts precipitate the abrupt consolidation of one outcome over others. The framework is then developed through three cases drawn from markedly different domains. The Arab Spring illustrates how authoritarian regimes persisted within Transition Zones, how legitimacy dissolved through threshold crossings, and how acts of observation — amplified through media and networked communication — reconfigured collective expectations into decisive outcomes. The 2008 financial crisis and the #MeToo movement extend the analysis to financial markets and the politics of gender, foregrounding in each a pivotal moment that accumulation accounts leave implicit: the constitutive act of observation by which a metastable situation resolves suddenly into one realized outcome. More broadly, QCM clarifies how uncertainty functions as a structural feature of social life rather than a deficit of information, reorienting social and political theory from prediction toward the navigation of potentiality. The framework carries implications for research methodology, governance, and policy — offering tools for identifying when systems are near threshold, when measurement becomes intervention, and when uncertainty is irreducible rather than merely incomplete. The formal framework is developed in a companion paper (Saviano 2025); the present article concentrates on its cross-domain application to episodes of discontinuous social and political change.

The framework of power convergence developed in prior work has shown how individuals can accumulate influence across structural, coercive, symbolic, psychological, and networked domains, operating beyond traditional institutional boundaries. Existing analyses, however, treat political authority as a downstream consequence of this process, leaving unresolved what occurs when convergence reaches the level of the state. This paper introduces the concept of recursive sovereignty to describe a threshold condition in which the relationship between the cumulative individual and the institutional order undergoes qualitative transformation. At this threshold, the individual no longer merely operates across domains of power but begins to constitute the conditions under which juridical authority is recognized and reproduced. Sovereign convergence is defined not by the scale of accumulated power but by a relational inversion in which institutional structures become integrated into the convergent system itself. The paper identifies three mechanisms through which this threshold is approached and crossed—legitimacy inversion, coercive alignment, and feedback closure—and examines them through comparative case studies of Nayib Bukele, Donald Trump, and Ibrahim Traoré. The cases are chosen to vary the strength of institutional buffering: Bukele represents rapid and legible crossing in a weakly buffered environment; Trump represents a staged and oscillating approach in a strongly buffered one, advanced through a second cycle whose outcome remains contested; and Traoré represents context-dependent crossing against a minimal institutional baseline. Together they demonstrate that the threshold is a relational condition defined by structural inversion rather than a fixed level of accumulation. A negative case—a fully convergent individual who does not cross the threshold—confirms that the concept discriminates between sovereign convergence and merely personalized or media-centered power. The analysis argues that sovereign convergence is characterized by a structural paradox: the recursive integration that produces short-term durability simultaneously generates systemic fragility, as tightly coupled domains raise the likelihood of cascading failure. By extending the framework of power convergence to the domain of sovereignty, the paper provides an analytical means of identifying when and how convergence may cross into a sovereign reconfiguration of political authority. NOTE: This paper is the third in a series on power convergence and the cumulative individual: Paper 1: A Theory of Power Convergence: The Cumulative Individual (January 2026) Paper 2: A Theory of Power Convergence: Epstein, Assange, and Ye as Cumulative Individuals (March 2026)

Recent years have witnessed the emergence of individuals whose influence appears to extend beyond the institutional positions they formally occupy. Actors operating outside traditional hierarchies have demonstrated the capacity to shape networks, narratives, and institutional responses across multiple domains of social power. Existing theories of power, however, typically explain influence through single institutional structures — states, markets, bureaucracies, or cultural systems — and therefore struggle to account for how individual actors can accumulate authority across these domains simultaneously. This paper applies the concept of power convergence and the figure of the cumulative individual, introduced in A Theory of Power Convergence: The Cumulative Individual (Saviano 2026), to a comparative analysis of three contemporary cases: Jeffrey Epstein, Julian Assange, and Ye (Kanye West). The framework proposes that individuals can acquire system-level influence when distinct domains of power — structural, coercive, symbolic, psychological, and networked — converge within a single actor. When these domains reinforce one another, individuals may operate as nodal points capable of mobilizing influence across institutional boundaries. The paper examines how each case represents a different pathway through which convergence can occur. Epstein's influence emerged through embedded elite networks and forms of infrastructural leverage within institutional systems. Assange's authority derived primarily from networked and symbolic power, while informational exposure also functioned as an indirect form of coercive leverage through the global dissemination of classified information. Ye's influence reflects the convergence of symbolic and psychological authority within global cultural media and digitally amplified communication networks. Taken together, these cases demonstrate how cumulative individuals can emerge through distinct configurations of convergent power domains. By comparing these cases, the paper argues that the cumulative individual represents an increasingly significant form of agency in highly networked societies. As communication systems, elite networks, and cultural economies become more tightly interconnected, the capacity for individuals to integrate multiple domains of influence may expand. Understanding the dynamics of power convergence is therefore essential for explaining how individual actors can shape institutional behavior and public discourse in contemporary social systems. A third paper in this series — Recursive Sovereignty: Power Convergence and the Threshold of Institutional Inversion (Saviano 2026) — extends the framework to sovereign-level convergence through comparative analysis of Nayib Bukele, Donald Trump, and Ibrahim Traoré.

This article develops a theory of power convergence centered on the concept of the cumulative individual. Existing theories of power remain analytically fragmented, isolating coercive, structural, symbolic, psychological, and networked dimensions without explaining how they interact. The theory advanced here argues that, under contemporary conditions of mediation, visibility, and connectivity, these domains increasingly converge within individual actors who function as systems of power rather than as occupants of discrete roles. Power is shown to accumulate through recursive feedback loops in which performance, amplification, affect, and institutional response mutually reinforce one another. To formalize this dynamic, the article introduces a developmental geometry of power organization and a systems model of convergence that specifies how heterogeneous domains become interoperable. The cumulative individual is defined as the entity in which convergent power stabilizes, reproduces itself, and remains inherently volatile. By providing a synthetic and extensible framework, the article establishes conceptual groundwork for analyzing emergent forms of authority that existing approaches cannot adequately explain. Subsequent work in this series applies the framework to empirical cases including Jeffrey Epstein, Julian Assange, Ye, Nayib Bukele, Donald Trump, and Ibrahim Traoré, and extends it to the analysis of sovereign-level power convergence and recursive sovereignty. AUTHOR'S NOTE: This working paper develops the core theoretical framework of a forthcoming book-length project, Power Convergence: From the Cumulative Individual to the Recursive Sovereign, which extends the model through empirical and comparative analysis. It is the first paper in a three-part series: Paper 2 applies the framework to Jeffrey Epstein, Julian Assange, and Ye as cumulative individuals operating through infrastructural, informational, and cultural convergence pathways; Paper 3 extends the framework to sovereign-level convergence, introducing the concept of recursive sovereignty through comparative analysis of Nayib Bukele, Donald Trump, and Ibrahim Traoré.

AI systems do not act. They generate coherent outputs under constraint. The persistence of agency attribution in artificial intelligence governance arises not from the presence of intention in these systems, but from a structural feature of their design: sufficiently constrained systems operating over structured data reliably produce the observable markers—coherence, responsiveness, apparent goal-directedness—that humans use to infer agency in other contexts. The inference is not irrational. It is misplaced. This paper identifies that misplacement as the central, undertheorized problem in AI governance. Across philosophy of mind, legal scholarship, and science and technology studies, AI systems are treated as candidates for agency—as full agents, quasi-agents, or distributed actants—despite the absence of the conditions agency requires. This paper argues that these frameworks share a category error: the conflation of output coherence with underlying subjectivity or purposive action. Drawing on philosophy of mind, legal theory, and science and technology studies, it develops a non-agentic model of AI systems as constraint-bound output generators and demonstrates how prevailing frameworks—the intentional stance, legal doctrines of responsibility, and distributed agency models—fail on their own terms when applied to systems of this kind. The governance consequences are direct. Misattributed agency produces unstable liability regimes, because responsibility cannot be correctly located when the wrong entity is treated as its source. It produces conceptual drift in policy language, as terms such as “AI decision-making” and “AI autonomy” migrate from heuristic to descriptive and embed false ontological assumptions in regulatory design. And it produces misplaced regulatory targets, directing institutional attention toward the system rather than toward the human decisions—in design, deployment, and use—where governance can actually take hold. This paper proposes a framework grounded in operational classification rather than metaphoric attribution. The reframing does not alter who bears responsibility—designers, deployers, and users remain accountable—but whether responsibility can be deflected. The agency frame grants AI systems enough ontological status to absorb causal attribution without bearing legal responsibility, creating a structural gap in accountability. The non-agentic model closes that gap: a constraint-bound output generator cannot function as a deflection point, and the defense—“the algorithm decided,” “the system determined”—becomes categorically unavailable. Every outcome traces to a human decision. That is what correct classification makes possible.

Research on conspiracy thinking has largely centered belief as its primary object of explanation, attributing conspiratorial beliefs to cognitive bias, psychological predisposition, or informational distortion. While this literature has generated important insights into individual-level mechanisms, it has struggled to account for the persistence, diffusion, and cross-context recurrence of conspiratorial belief — patterns that suggest the presence of upstream structural conditions. This article argues that these limitations stem from a mis-specified analytic focus: belief is treated as a causal variable rather than as an outcome shaped by institutional conditions. Drawing on a selective review of psychological, sociological, and media-oriented approaches, the article identifies a shared assumption underlying belief-centered models — that conspiratorial belief reflects epistemic deficiency rather than rational adaptation to institutional environments characterized by opacity, weak accountability, and narrative insufficiency. Under such conditions, belief correction alone cannot stabilize authoritative interpretation, and debunking interventions consistently underperform. Rather than proposing a comprehensive theory, the article offers a reframing intervention that shifts analytic attention from belief content to the institutional conditions that structure credibility, epistemic authority, and interpretive stability. This shift has implications for research design, measurement, and comparative analysis — and for understanding why institutional trust erodes even in the absence of deliberate deception. This working paper develops the conceptual framework expanded and applied in Legitimate Distrust: Why Conspiracy Theories Grow When Institutions Fail (Statera Press, 2026), Book Two of The Collapse of Trust series.

The philosophy of social science faces a persistent explanatory gap in reflexive systems: when modeling alters what is being modeled, predictive frameworks become unstable at the moments they are most needed. Complexity theory and reflexivity each address part of this problem — nonlinearity, feedback, recursive expectation — but each leaves comparatively underspecified the moment at which structured indeterminacy resolves into a realized social outcome. This article proposes Potentialism and the Quantum-Continuity Model (QCM) as a candidate epistemological framework. Rather than claiming social systems are literally quantum, QCM treats quantum theory as a source for disciplined structural analogy, transferring the bare structure by which multiple live possibilities resolve into a definite outcome — while explicitly bounding what does and does not carry over. From this it generates three concepts: Transition Zones, Quantum Thresholds, and Quantum Measurement Events. The framework positions itself among existing accounts of constitutive observation — self-fulfilling prophecy, performativity, and second-order systems theory — specifying what it adds to them. The Arab Spring serves as illustration, with briefer extensions to the 2008 financial crisis and #MeToo. QCM offers a vocabulary for discontinuous social resolution, reorienting inquiry from prediction toward the navigation of structured indeterminacy.

Moral progress is not one movement but two. The first raises the floor: it deepens what counts as an impermissible harm. The second expands the radius: it widens the class of persons to whom the floor is owed. Classical Rome protected the freeborn child against sexual violation with real legal force, but the enslaved child of the same age, in the same act, fell outside protection entirely. Extending protection to the enslaved child would have widened the radius without raising the floor. Counting subtler harms against freeborn children as wrongs of the same order would have raised the floor without widening the radius. Different reforms, different beneficiaries, different moral recognitions. These two movements are logically independent, they have usually been run together, and the conflation disarms the progressive narrative against the relativist it most wants to refute: if the floor just is the leading edge of the radius, there is no fixed moral object the radius is moving toward. Holding the floor and the radius apart restores the stable moral object the conflation removed — but only on two separated tracks. On a reality-track, three justificatory traditions selected for the substantial differences in their moral vocabularies — Roman status-honor, the Hebrew covenantal-pollution register, canon-law infantia — converge on the same protected boundary, warranted by the Shklarian priority of injury over principle. On an authority-track, the Street–Joyce debunking objection is confined rather than defeated. Causal explanation cannot, by itself, settle justificatory content for those already engaged in giving and asking for reasons. The genealogy explains why we reason morally at all; it does not determine what our reasoning concludes. The two tracks are analytically distinct: advances on one dimension do not entail advances on the other. History can establish the object; it cannot establish the object’s authority. The paper does not pretend to deliver both from one source. The keystone is that the moral floor is a doctrine about victims, not agents, grounded in vulnerability rather than rational capacity. The consequence is that Kantian content (the Formula of Humanity) can be recruited without Kantian grounding. The framework is deliberately hybrid. Its individual components have clear precedents in the literature, but their combination yields a distinctive argumentative result.