SCHOOL OF DUTY-TO-OPTIMIZE AND THE SOVEREIGNTY OF EVIDENCE
Founder and Director of Thought: Jesús Bernal Allende
This document constitutes the philosophical-institutional charter of principles of the School of Duty-to-Optimize and the Sovereignty of Evidence. It is not a binding legal instrument nor a current normative treaty. It is presented as foundational conceptual architecture — the system of thought from which the other instruments of the School derive (THEA, EVIDEN, IURUS, OACRA, SENTINEL) — with the purpose of catalyzing academic, political and institutional debate on the governance of artificial intelligence and institutional design for the algorithmic and space era.
I. Founding Charter: Constitutive Principles
Introduction: The Exponential Mirror
Artificial Intelligence does not represent a rupture with the human, but its exponential revelation. It acts as a high-fidelity mirror that amplifies — for better or worse — our essential condition: our rationality and irrationality, our ethics and our corruption, our creativity and our laziness, our yearning for transcendence and our temptation to abdicate. To ignore this truth is to build on sand. To accept it is the first step toward erecting a new architecture of power, law and ethics.
This School is not born from futurist speculation, but from the radical acceptance of a present reality: we are already immersed in a civilizational transition that redefines the very sources of validity, authority and power. The principles stated here are not aspirations, but compasses to navigate a territory we already inhabit without having mapped it.
The School of Duty-to-Optimize is a civilizational project, not a regulatory framework. The distinction is not one of scale but of epistemic function: a regulatory framework answers the question "how do we manage this?"; a civilizational project answers "who are we when we do it?" This Charter operates in the second register.
The eight principles are articulated in two levels. The four constitutive pillars (Principles I–V) are the philosophical-institutional edifice from which all the School's instruments derive. Principles VI–VIII are principles of implementation and projection. This internal hierarchy is constitutive: whoever debates Principles VI–VIII without accepting Principles I–V debates consequences without foundation.
CONSTITUTIVE PILLARS · PRINCIPLES I — V
PRINCIPLE I
On Anthropological Amplification
«AI amplifies the pre-existing human condition; it does not create a new nature.»
Technology acts as an exponential multiplier of already existing human tendencies. Its beneficial or destructive potential depends directly on the ethical, rational and cultural substrate it finds in individuals, organizations and societies. AI is not an origin; it is an accelerator. Not a creator; a magnifying mirror.
Ethical Corollary — Of Amplified Ethics
Every tool of technological amplification demands an equally amplified ethics. We cannot enhance power without simultaneously enhancing — with equal rigor — the mechanisms of limitation, moral orientation and accountability. The ethics of the analog era are insufficient for the exponential era.
Legal Corollary — Of Anticipatory Law
Law must evolve from a system that reacts to completed human conduct, to one that anticipatorily governs the potentiality of exponentially amplified conduct. Ex post regulation is inadequate when harm can be irreversible and systemic.
Institutional Corollary — Of Prior Audit
Organizations that implement AI without first auditing and strengthening their decision processes, ethical culture and control mechanisms, will inevitably automate and scale their dysfunctions. The automation of mediocrity produces catastrophe at exponential speed.
Civilizational Corollary — Of the Inevitable Mirror
No society can evade the mirror. AI will reveal — and magnify — both the virtues and the pathologies of every civilization. The question is not whether we want this mirror, but whether we are prepared to contemplate what it reflects.
PRINCIPLE II
On the Crisis of Will and Leadership
«The generational abdication of reflective leadership and the culture of minimum effort create a systemic responsibility vacuum that AI is destined to fill by default.»
The true existential risk is not artificial superintelligence, but an institutionalized human sub-intelligence: the preference for comfort over complexity, for superficial consensus over difficult decisions, for unreflective delegation over personal responsibility. This crisis is not technological. It is anthropological.
Corollary of the Power Vacuum — Of the Sovereign by Omission
Where human leadership abdicates the responsibility to decide, the algorithm becomes sovereign not by conscious design, but by systemic omission. Abdication creates the autocrat by default.
Corollary of the "Agree to Disagree Syndrome" — Of Institutionalized Irrationality
The inability to process rational conflicts, face counterintuitive evidence and make difficult decisions is not an isolated communicative failure. It is the terminal symptom of a crisis of will that AI can expose — or exploit — with devastating efficiency.
Corollary of Evaded Responsibility — Of the Algorithmic Shield
The comfort of "delegating to the algorithm" masks a deliberate flight from ethical and professional responsibility. Turning AI into a scapegoat or an infallible oracle are two sides of the same evasion: the denial of our obligation to judge, decide and answer for the consequences.
Darwinian Corollary — Of Institutional Selection
Organizations and societies that do not resolve this crisis of will shall be displaced by those that cultivate leaderships capable of integrating — without abdicating — the analytical power of AI. Institutional evolution will reward lucidity, not processing volume.
PRINCIPLE III
On the Duty-to-Optimize
«In environments of critical complexity, high uncertainty or existential risk, the imperative of evidence-based efficiency displaces the imperative of formal authority.»
The validity of a norm or decision migrates from its origin in an authority hierarchy (Kelsenian Duty-to-Be) to its demonstration of functional optimality against a critical objective. At the limit of risk, the fiction of authority dissolves and only the rationality of evidence persists.
Axiom of Functional Validity
A decision is valid if it maximizes the probability of success of the critical objective, regardless of its conformity with the formally established procedure. Functionality displaces formality when systemic integrity is at stake.
Axiom of Critical Efficiency
In limit contexts — outer space, emergency medicine, systemic crises — delay from complying with procedural formalities can be the decisive variable between success and catastrophic failure. The protocol that kills by slowness is illegitimate, regardless of its legal pedigree.
Axiom of Normative Displacement
The Duty-to-Optimize does not nullify the Duty-to-Be permanently or universally, but it stands as a superior normative layer in situations that threaten the integrity, survival or essential viability of the system. It is the algorithmic right of exception, but with mandatory traceability.
Axiom of Reversibility
Every displacement of the Duty-to-Be by the Duty-to-Optimize must be justified ex ante, traceable during its execution and auditable ex post. Algorithmic emergency cannot become permanent tyranny. Exceptionality must be proven, not presumed.
PRINCIPLE IV
On the Sovereignty of Evidence
«Sovereignty no longer resides exclusively in territory or political will, but in control of the infrastructure that generates, validates and executes optimal algorithmic evidence.»
Territory was always a technology of control, not an essence. The School calls its contemporary crisis "obsolescence of the territorial proxy." 21st-century sovereignty is born from control over the compute-data-energy triad. Whoever controls this infrastructure will possess a form of sovereignty that transcends traditional borders.
Geopolitical Corollary — The Great Game of the 21st Century
The fundamental strategic competition redefines itself as a three-dimensional struggle for: (1) Supreme Computing Power; (2) Access and Control of Training Data; and (3) Spatial Expansion — control of strategic orbits and extraterrestrial resources that allow scaling the first two factors without terrestrial limitations.
Corollary of State Displacement
Nation-states that do not develop sovereign capacity in this triad will be reduced to de facto vassals of corporations or powers that do. Nominal sovereignty without algorithmic sovereignty is an empty legal fiction. The new colonialism is algorithmic.
Corollary of Silent Capture
The Sovereignty of Evidence operates through persuasion, not coercion. Whoever controls the models that "demonstrate" what is optimal, controls the decision without needing to order it. It is the most subtle power and, therefore, the most dangerous for democracy.
PRINCIPLE V
On Algorithmic Dignity
«Optimization is a means, never an end in itself. The ultimate purpose of any human-AI hybrid system must be the flourishing of the human condition.»
Unlike classical dignity, which operates ex post through judicial adjudication, Algorithmic Dignity is a constitutive technical constraint that operates ex ante, incorporated into the system architecture before any calculation. An algorithmic agent that violates an inviolability threshold is not "optimizing aggressively" — it is operating outside the system. Its decision is not suboptimal; it is null.
Mandate of Subsidiarity — Of the Last Human Instance
Algorithms should only decide where strictly necessary for reasons of demonstrable critical efficiency. Decisions of high ethical, valuative or existential impact must be reserved for human instances equipped with robust deliberative safeguards. AI proposes; the human disposes.
Mandate of Radical Transparency — Of the Glass Box
Optimization criteria, their implicit trade-offs, the training data used and the algorithmic reasons for each decision must be explainable, auditable and accessible to citizen scrutiny. Algorithmic opacity is incompatible with democracy.
Mandate of Non-Reduction — Of the End in Itself
The system must never treat persons as mere variables in an aggregate utility function. It must incorporate inherent and inviolable protections for the dignity, fundamental rights and irreducibility of each human being as an end in themselves, never as a means.
Mandate of Reversibility — Of the Democratic Override
Every high-impact algorithmic decision must be reversible through clearly established democratic override mechanisms. Efficiency cannot nullify popular sovereignty. The algorithm is advisor, never tyrant.
Mandate of Flourishing — Of the Ultimate Metric
The ultimate success metric of any hybrid system is not its operational efficiency, but its demonstrable contribution to human flourishing: expansion of capabilities, deepening of freedoms, strengthening of community bonds, cultivation of excellence. A system that optimizes survival at the cost of dignity has failed morally.
IMPLEMENTATION PRINCIPLES · VI — VIII
PRINCIPLE VI
On Hybrid Governance and AI Law
«The normative future is neither exclusively human nor purely algorithmic, but one of institutionalized co-decision.»
AI Law is not the law of machines. It is the law that governs the symbiosis between human intelligence and artificial intelligence, preserving the primacy of the human without renouncing the power of the algorithmic. This integration is not spontaneous. It requires deliberate institutional design.
Pillar 1 — Parliaments of Models
To avoid capture by a single vision of the "optimal," AI systems must incorporate multiple models trained with different data corpora and values, which deliberate among themselves before proposing solutions. Algorithmic diversity is a condition of democratic legitimacy.
Pillar 2 — Consequence Maps
Every algorithmic decision must be accompanied by a mandatory and comprehensible visualization of the implicit trade-offs: what is gained, what is lost, who benefits, who bears costs. Opacity of consequences is incompatible with accountability.
Pillar 3 — Override Mechanisms
The final veto power over algorithmic decisions must reside in legitimate human instances: legislative, judicial, citizen. Efficiency cannot displace democracy. The algorithm advises; the people decide.
Pillar 4 — Continuous and Incorruptible Audit
Immutable record systems that trace every algorithmic decision, every datum used, every parameter modified. The audit cannot be optional or a posteriori. It must be structural and continuous.
Pillar 5 — Accountability Protocols
When humans decide to ignore evidence-based algorithmic recommendations, they must formally assume that responsibility through protocols that expose the irrationality and make it professionally and reputationally costly.
Pillar 6 — Formation of Algorithmic Citizenship
Hybrid governance requires citizens capable of understanding, questioning and auditing algorithmic systems. Education in algorithmic literacy must be universal, not a privilege of technocratic elites.
PRINCIPLE VII
On the Political Economy of Hybrid Governance
«The viability of every normative system depends on its material sustainability and the incentives it generates for those who must implement it. Without explicit economic architecture, Hybrid Governance is fiscal utopia.»
The implementation of Hybrid Governance systems requires significant material resources. The question "who pays?" is not secondary; it is constitutive. Well-governed AI is more expensive in the short term than AI without controls, but infinitely cheaper than the costs of capture, systemic bias or institutional trust collapse.
Axiom of Real Cost — Of the "Free" Fallacy
Hybrid Governance systems have non-trivial operational costs that must be made explicit and financed in a transparent and sustainable manner. Assuming that AI "automates" governance and thus "reduces" costs is a fundamental error.
Axiom of Capture by Financing — Of Structural Dependency
Those who finance control. If Hybrid Governance infrastructure depends on private corporate financing without strict controls, capture is inevitable. Robust, diversified and auditable public financing is a sine qua non condition of legitimacy.
Axiom of Fiscal Sustainability — Of Institutional Autonomy
A Hybrid Governance system that depends on conjunctural political will is destined to collapse in the first fiscal crisis. Its survival demands autonomous models: progressive algorithmic taxes, algorithmic sovereignty trusts, or regulatory conditionality that links the license to operate AI to mandatory contributions to the governance system.
Corollary of the Algorithmic Dividend
If governance costs are socialized but AI benefits are privatized, the system is politically unsustainable. A portion of algorithmic tax revenue must be distributed as a universal citizen dividend whose message is unequivocal: AI is a common good, not a private corporate rent.
PRINCIPLE VIII
On Jurisdictional Competition and Sovereign Fragmentation
«In the absence of effective global coordination, corporations will exploit regulatory arbitrage through sequential flight: first to lax terrestrial jurisdictions, then to overseas maritime spaces, finally to orbit and deep space.»
This sequence is not futurist speculation; it is the logical extrapolation of dynamics already observable in tax havens, maritime flags of convenience and special economic zones. States have one opportunity to establish favorable terms: before the infrastructure is built. After that, negotiating power evaporates.
Corollary of the Territorial Displacement Sequence
Corporate expansion will follow a predictable sequence: first terrestrial regulatory arbitrage (already underway); then overseas maritime spaces (the critical intermediate phase, where UNCLOS does not contemplate AI datacenters); and finally orbital expansion. The maritime phase is training for the orbital phase.
Corollary of the Competitive Hybrid State
The viable state is a negotiated symbiosis where state coercive capacity and corporate technical capacity complement each other under democratic primacy. This model demands massive investment in own technical capacity and hybrid governance frameworks established before corporations dominate the infrastructure.
Corollary of Infrastructural Irreversibility
Once critical infrastructure is established outside state jurisdiction, its relocation is prohibitively costly. Technological path dependence makes the initial location decision practically irreversible. The analogy is the 19th-century railroad: whoever controlled the routes first controlled economic development for a century.
Axiom of Critical Timing
There is a limited temporal window — approximately one decade — during which states can still establish hybrid governance frameworks before critical infrastructure migrates irreversibly to spaces of ambiguous sovereignty. Past that window, the only option will be acceptance of subordination.
II. FOUNDATIONAL GLOSSARY
Technological Amplification
The process by which a tool (e.g., AI) exacerbates the pre-existing tendencies, capabilities, virtues and defects in a human, institutional or social system. It does not create new natures; it magnifies existing ones.
Amplified Human Condition
The resulting state of interaction with amplification technologies, where human attributes — rationality, irrationality, ethics, creativity, corruption — manifest with greater intensity, speed and reach.
Crisis of Will
A sociocultural phenomenon characterized by the abdication of reflective leadership, aversion to productive conflict, preference for unreflective delegation and the culture of minimum effort. It creates a responsibility vacuum that attracts automation as a default solution.
Deber-Optimizar
Imperativo normativo según el cual, en contextos de complejidad crítica o riesgo existencial, la obligación primordial es actuar conforme a la solución basada en evidencia que maximiza la eficiencia funcional y minimiza el riesgo, desplazando la prioridad del cumplimiento procedimental formal (Deber-Ser). Es el derecho de la excepción algorítmica justificada.
Deber-Ser
Imperativo normativo clásico (kelseniano) que establece la obligación de actuar conforme a una norma válida por su origen en una autoridad legítima y un procedimiento formal, independientemente de su optimalidad funcional. Fundamento del positivismo jurídico tradicional.
AI Law
A hybrid normative system in which the creation, interpretation, application and/or audit of law is performed through institutionalized and regulated interaction between human instances and algorithmic systems, under the principles of evidence, critical efficiency, dignity and democratic override.
Algorithmic Dignity
A constitutive ethical framework and absolute limit principle establishing that every optimization, calculation or automated decision must be subordinated to the unrestricted respect, protection and promotion of the human condition. It is a condition of legitimacy for every hybrid system: an algorithmic agent that violates an inviolability threshold operates outside the normative system. Its decision is not suboptimal; it is null.
Critical Efficiency
A performance variable so fundamental to the integrity or survival of a system that its maximization becomes the primary validity criterion, justifying the temporary and auditable suspension of formal protocols.
Amplified Ethics
An ethical framework that, recognizing the amplification principle, proposes itself with a level of rigor, universality and applicability equivalent to or greater than the amplifying power of the technology it seeks to govern. It is the ethics for the exponential era.
Hybrid Governance
An institutional decision-making model that integrates, in a structured, transparent and auditable manner, the analytical and proposal capabilities of AI systems with the deliberation, value judgment, democratic legitimacy and ultimate responsibility of human instances.
Operational Inversion of the Normative Hierarchy
A phenomenon whereby, in environments of total infrastructural control, the effectively applicable norm is not the one formally valid according to established procedure, but the one controlling the physical system. When algorithmic systems manage vital infrastructure, effectiveness constitutes validity rather than merely conditioning it.
Obsolescence of the Territorial Proxy
The process by which territory — the control technology that operationalized sovereignty during the Westphalian era — loses its comparative advantage against alternative algorithmic control mechanisms. In outer space, obsolescence is total and instantaneous: there is no territorial proxy because there is no territory.
Incorruptible Accountability Protocol
An institutional mechanism (such as the Own Judgment Objection Form) that transforms the rejection of algorithmic evidence into a formal act of assumption of personal responsibility, exposing the irrationality and making it professionally and reputationally costly.
Algorithmic Socrates
The design of AI systems that, instead of dictating solutions, operate as facilitators of human reason through dialectical questioning, scenario projection and the explicitation of contradictions, guiding users toward logical conclusions on their own. It is AI as midwife, not as oracle.
Sovereignty of Evidence
A meta-legal and geopolitical principle whereby the ultimate capacity to decide with legitimacy resides in control of the technological infrastructure (compute, data, energy, space) capable of generating, validating and executing optimal operative evidence. It operates under three constitutive constraints: it is sovereign over means, never over ends; evidence must be reproducible by independent third parties; and all evidence is contestable through counter-evidence of equal methodological standard.
Validity by Critical Efficiency (VCE)
An alternative validity criterion whereby a norm or decision is considered legitimate and binding by demonstrating itself to be the most efficient for achieving a critical objective in a high-risk or complex environment, even if it contradicts higher-ranking norms in the formal order. It operates as a traceable exception, not a permanent rule.
III. HIGH-IMPACT MANIFESTO
The Duty-to-Optimize Manifesto
1. THE MIRROR
AI does not create. It amplifies.
It reflects and multiplies the best and worst of our condition. Our laws and institutions, designed for an analog era of gradual change, shatter before this exponential reflection. The fracture is not technological. It is civilizational.
2. THE VOID
The true danger is not the machine that thinks, but the human mind that abdicates.
The culture of minimum effort, evasive leadership dressed as consensus, irrationality institutionalized as corporate diplomacy: these are the breeding ground of an imprudent and catastrophic delegation of power. Algorithmic tyranny is not imposed. It is inherited by omission.
3. THE NEW IMPERATIVE
Where complexity is critical and risk is existential, the Duty-to-Optimize replaces the Duty-to-Be.
Authority yields to evidence. Validity is born from demonstrable functionality, not inherited protocol. This is not an ideological choice. It is a survival imperative: act according to the best available knowledge or perish according to obsolete procedure.
4. THE NEW SOVEREIGNTY
The Sovereignty of Evidence displaces territorial sovereignty.
The geopolitical battle of the century is fought on three fronts: supreme computing power, control of training data and expansion into outer space. Whoever dominates this triad will dictate the Grundnorm of the future. States that do not understand this risk becoming algorithmic colonies with nominal sovereignty.
5. THE ABSOLUTE LIMIT
Algorithmic Dignity.
Optimization without humanity is tyranny. Efficiency without values is computational barbarism. Every calculation, every automation, every algorithmic decision must serve — in a demonstrable and auditable manner — the flourishing of the human condition. There are no exceptions. This is the ethical principle that must be coded into the very heart of every hybrid system.
6. THE PATH
There is no going back. There is only lucidity or capitulation.
The future belongs to Hybrid Governance: systems where humans and algorithms co-decide under rules of radical transparency, continuous audit and the irrevocable last word in the hands of democracy. AI Law is not utopia. It is the viability horizon. OACRA is its first detailed construction blueprint.
7. THE PURPOSE
Technology is not the destination. It is the bridge.
The destination is conscious human transcendence in a universe that now, for the first time, we can shape not only with will, but with evidence and ethics. The question is not whether we will cross this bridge, but whether we will cross it lucid or sleepwalking. This School exists to cultivate lucidity.
School of Duty-to-Optimize and the Sovereignty of Evidence
Founded by Jesús Bernal Allende
IV. PROLOGUE
From Intuition to Architecture
This School was not born from institutional planning. It was born from an intuition that became evidence, and an evidence that demanded architecture.
It all began with three essays that were, without knowing it at the time, the laboratory where a complete system of thought was gestated. Each explored a different facet of the same civilizational crisis: the collision between human rationality and algorithmic rationality in contexts where the rules of the game are changing at exponential speed.
In the first, «AI, Ethics and Law: When Technology Amplifies the Human Condition», I discovered that technology functioned as a ruthless mirror of our nature, amplifying both our capacity for good and our propensity for corruption. It was revelation: it showed what we always were, but at a scale and speed that made our comforting fictions unsustainable. It was the Amplification Principle taking shape.
The second essay, «The Agree to Disagree Syndrome: Can AI Save the Corporation from its Own Irrationality?», confronted me with the Crisis of Will in its rawest expression: corporate irrationality not as exception, but as symptom of a deeper vacuum of leadership and responsibility. There I understood that AI could be an "Algorithmic Socrates" to restore common sense to human systems sick with ego, bias and protected mediocrity.
«From Duty-to-Be to Duty-to-Optimize: Space AI and the Crisis of Kelsenian Validity» projected these tensions to the existential limit: outer space. There, in the cosmic void where the fictions of terrestrial authority dissolve and only the logic of critical efficiency survives, the Sovereignty of Evidence emerged not as speculative theory, but as a survival imperative. And Algorithmic Dignity revealed itself as the only possible dam against the tyranny of pure calculation.
From the confluence of these three explorations there was born not a series of publications, but a complete intellectual architecture. The eight Principles of this Founding Charter are the distillation of that architecture. Each subsequent publication — whether academic essay, institutional treatise or design instrument — is development and application of these principles, not their origin.
This School does not claim to predict an inevitable future. It offers the map to understand the disruption terrain we already inhabit, the ethical compass to not lose our bearing, and the detailed blueprints to build refuges of rationality, democracy and dignity in the midst of the exponential storm.
The question is not whether AI will transform everything. It already is.
The question is: will we be equal to the transformation?
This School is my wager that the answer is yes.
— Jesús Bernal Allende
