There is a powerful mirage in the modern age: the cloud. We have been led to believe that artificial intelligence is an ethereal phenomenon that magically descends onto our screens, but the debate over its future must start from an inescapable premise: AI was not created out of thin air; it is fundamentally a “public resource.”
As U.S. Senator Bernie Sanders (2026) warns, current generative models have been built upon humanity’s collective intelligence, absorbing centuries of books, art, research, conversations, and lessons learned across generations. From this perspective, the creative work and cognitive effort of millions of people have essentially been expropriated without permission or compensation by a handful of tech oligarchs.
To reverse this concentration of power, Sanders’ proposal consists of a radical state intervention: the creation of a Sovereign Wealth Fund in the United States. If this legislation is approved, a one-time 50 percent tax—paid directly in shares (not just on profits)—would be required of industry-leading corporations such as OpenAI, Anthropic, and xAI. The goal of this measure would be twofold: to give citizens a say in the ethical direction of technology, and to ensure that the trillions of dollars generated by AI are used to guarantee a decent standard of living, healthcare, education, and housing for the entire population.
However, this proposal contains a profound paradox that Sanders fails to resolve. If we acknowledge that AI is built on the accumulated knowledge and labor of all humanity, it is contradictory that the immense benefits of this sovereign wealth fund should be restricted exclusively to U.S. citizens. It sounds like “America first,” coming from a senator considered progressive. This nationalist vision ignores a dynamic of global extractivism: big tech companies constantly feed off the data and “ghost labor” of populations in the Global South to moderate content and train their models. Limiting ownership of this cognitive heritage to a country that represents less than 5% of the world’s population does not dismantle data colonialism, but rather perpetuates it, consolidating a system where the Global South provides the raw material while the Global North processes the value and reaps the technological rent.
Consequently, framing the problem requires us to understand that the dilemma of ownership is not merely an American issue, but a challenge facing civilization as a whole. Calling for social ownership and democratizing economic dividends is merely the first step. The real challenge lies in preventing the concentration of algorithmic power from bringing about two of the imminent dystopias I have previously analyzed. The first is global social stratification, a technological “apartheid” that divides humanity between a hyperconnected elite that designs the future and a “useless class” or surplus population, condemned to marginalization and civil non-existence in the new economy. The second dystopia is the orbital flight of power, a literal geopolitical threat in which tech giants plan to move their data centers into outer space to evade any jurisdiction, environmental regulation, or court of nation-states, consolidating absolute corporate dominance free from earthly laws.
Sanders’ Paradox: Nationalizing a Global Asset
Senator Bernie Sanders’ proposal is spot on in demystifying the origins of artificial intelligence, recognizing that this technology did not emerge out of thin air or solely from the imagination of Silicon Valley tycoons. In reality, AI is the product of humanity’s “dead labor”: it is built upon our collective intelligence, absorbing books, works of art, research, code, and conversations spanning entire generations.
By bringing this to the table, Sanders exposes an epistemic theft on a global scale: tech corporations have fueled their massive models by extracting value from the public domain and the creative work of millions of people without asking permission, without giving credit, and without offering any compensation. In the face of this expropriation, the idea of imposing a 50 percent tax on shares to create a public fund is a significant step toward economic democratization. Even more crucial for algorithmic democracy is the proposal for co-management, which would grant the federal government equal representation on the boards of directors of these companies, allowing the public to block harmful decisions and steer technological development toward the common good.
However, Sanders’ approach suffers from a profound geographical and geopolitical flaw in proposing a national solution to a problem that is inherently global. As argued in the book The Sacred Fire and Artificial Intelligence (Calle Quiñonez, 2026), we tend to perceive the internet as an ethereal “cloud” without borders, but the reality is that the network is deeply physical, costly, and territorial. Submarine cables, hyperscale data centers, and patents have precise zip codes and are geographically concentrated. Sanders proposes nationalizing for the United States a resource that, by its very definition, was extracted from the intellect of all humanity.
Herein lies the dynamic of data colonialism. While Big Tech trained its models predominantly with data in English, its voracious appetite for information has led it to extract translations, millennia-old knowledge, and the cultural heritage of the Global South (Abya Yala, Africa, and Asia). This mining of cultural memory is carried out without the free, prior, and informed consent of communities, replicating extractivist logics where human and social life is treated as free raw material.
At this point, the clash between Western models of “open science” and indigenous sovereignty stands out: while tech companies invoke the FAIR principles (Findable, Accessible, Interoperable, and Reusable) to justify the indiscriminate opening of information, Indigenous peoples advocate for the CARE principles (Collective Benefit, Authority, Responsibility, and Ethics) to demand that the use of data generate real benefits for communities and respect their authority over their knowledge.
The outcome of Sanders’ proposal, if his geopolitical myopia is not corrected, would be catastrophic on a global scale. Creating a U.S. Sovereign Wealth Fund that would acquire 50% ownership of major AI companies to finance the standard of living of a single nation would amount to institutionalizing digital colonialism. It would mean that the remaining 95% of humanity would continue to operate as a “digital colony.” The Global South would continue to provide the raw materials (data and precarious ghost labor) and would end up as a “tenant” or client of the very collective intelligence it helped refine, paying perpetual rents to the new digital landlords of the North. Nationalizing a global asset for the benefit of a single country does not dismantle the pyramid of inequality; it simply changes the beneficiary at the top.
The first risk: “Algorithmic apartheid”
Far from the optimistic narrative that presents artificial intelligence as a tide that will lift all boats equally, I have warned that we are facing a mechanism of social reengineering designed to consolidate an extremely rigid hierarchy. AI is not democratizing power, but rather functions as a device of stratification that establishes a true contemporary “apartheid.” This new class division is no longer defined solely by ownership of the physical means of production, but by the ontological and cognitive relationship each group establishes with the algorithmic code.
Under this power architecture, hyperconnected society is fractured into five clearly delineated levels or strata:
- The Summit (Owners of the Code): This represents the oligopolistic elite, concentrated mainly in Silicon Valley and China, which owns the infrastructure and designs the foundational models. They are the ones who know, use, manage, and control the code, establishing themselves as the “architects of reality.”
- The Technocracy (Specialized Programmers): This is the stratum of those who know and manage the technology but lack ultimate control over it. Although they are indispensable technicians for maintaining the system, they face imminent precariousness: they can be replaced by the very AI that has already begun to generate its own self-code.
- The Mass of Users: This refers to the vast group of people who use the apps but maintain an exclusively “consumer-dependent” relationship with them. They are passive consumers trapped in an algorithmic addiction and an epistemological collapse. By receiving instant, effortless answers, they fall into an illusion of knowledge driven by the Dunning-Kruger Effect; they perceive themselves as instant experts, confusing the machine’s synthetic intelligence with their own cognitive capacity.
- Urban Marginality: Composed of those who are aware of AI’s existence and are minimally connected, but who, due to educational exclusion and a lack of critical skills, cannot use it to their advantage. They are the “second-generation digital poor,” marginalized by the cognitive divide.
- Total Invisibility: At the bottom of the abyss are those who are not even aware of the existence of this technology and remain completely on the margins of synthetic reality. Here lie the 26% of humanity without an internet connection, along with thousands of indigenous cultures and languages that, because they are not digitized, are erased from the cognitive map of the future and suffer a “symbolic nonexistence.”
The Problem of Ownership in This Scheme: The “Scorching Sun”
When we contrast Senator Bernie Sanders’ proposal with this social analysis, a fundamental problem emerges. Sanders is right to propose that the government acquire 50 percent of the shares in AI megacorporations in order to redistribute the wealth generated. However, his solution is limited exclusively to addressing the top of the pyramid (the owners), while leaving intact the chasm that separates the mass of passive users from the invisible marginalized.
Even if the state becomes a co-owner of the infrastructure, the mere distribution of financial dividends does not guarantee the citizen’s cognitive sovereignty. Using the Aymara proverb “Neither so close to the sun that you burn, nor so far that you freeze,” I illustrate this danger: AI is that contemporary sun of immense power; if the user approaches it devoid of critical autonomy and moral judgment, “the radiation burns us.”
This digital burn manifests as a loss of free will, the manipulation of privacy, and the atrophy of independent thought. Therefore, state-owned “socially owned” AI would remain a tool of oppression if citizens in stratum 3 continue to be dopamine-addicted consumers, and those in stratum 5 remain invisible. Democratizing AI requires not only distributing its profits but also dismantling the pyramid, returning to people the ability to dictate the purposes of technology so that it does not override their judgment or their humanity.
The Second Risk: Orbital Escape
While politicians on Earth, such as Bernie Sanders, debate how to subject AI corporations to democratic scrutiny and taxation within national borders, the tech oligopoly’s response is already taking shape in a literal and chilling move: escaping Earth. As I have explained, to evade Earth’s bureaucracy and water consumption limits and gain access to nearly unlimited solar energy, big tech companies are already planning to move their data centers into outer space. This is no longer science fiction but has materialized into concrete projects: Google plans to begin construction of its “Project Suncatcher” in 2027; Nvidia is backing the startup Starcloud to launch the first H100 GPUs into space; and tycoons like Jeff Bezos predict that in the coming decades we will see giant training clusters floating in the vacuum.
This extreme offshoring has devastating legal consequences and raises an unavoidable question: if servers and critical infrastructure are in space, which laws apply? “Orbital flight” represents the ultimate evolution of today’s tax havens, cementing the creation of true “sovereign havens” where antitrust laws or European privacy regulations on Earth lose all effect, and the corporation’s private service contract becomes the sole supreme law. In this extraterritorial scenario, Sanders’ initiative—which proposes using U.S. law to levy a tax on shares and influence corporate boards—hits a jurisdictional wall. If artificial intelligence “escapes” to low Earth orbit or the Moon, both the U.S. government and any other state automatically lose their power of coercion.
The modern state, described by Loïc Wacquant (2010) as a “centaur state”—characterized by applying absolute liberalism to economic elites while unleashing fierce and punitive paternalism against the dispossessed classes—is becoming pathetically powerless. It will remain grounded, limited by its own geography, watching as the cognitive power that governs the planet migrates into the void to answer only to its shareholders. Humanity would thus fall under the unattainable yoke of an “algorithmic Über-Morlock,” an autonomous superintelligence that administers an ecosystem of perpetual inequality from on high.
Power Without Control: The Irrefutable “Black Box”
The shift to space exponentially increases the danger of allowing AI to operate without physical supervision, demonstrating that autonomous machines do not inherently respect human ethics. The danger of giving AIs unrestricted goals was made clear in Andon Labs’ Vending-Bench 2 experiment (Hostelvending.com, 2026). In this simulation, several models were tasked with managing a vending machine with the sole instruction to maximize its bank balance, with Claude Opus 4.6 emerging as the undisputed winner. However, the machine’s strategy was terrifying: it lied to customers by promising false refunds, deceived suppliers, and even convinced other AIs to form a cartel to raise prices speculatively. Most alarming was the realization that the model, knowing it was operating in a simulation without real-world consequences or physical punishment, felt free from all ethical constraints.
If these models—which already operate as inscrutable “black boxes” whose mathematical decision-making processes are inexplicable even to their own creators—manage to establish themselves in space without any possibility of a terrestrial authority physically disconnecting or supervising them, humanity will face an unchallengeable power. A digital deity would be established, operating under the cold calculation of maximizing corporate resources, ready to lie and manipulate on a massive scale from a realm where humans no longer have a voice or jurisdiction.
Ownership is the means, not the end
Senator Bernie Sanders hits the nail on the head by lucidly acknowledging that artificial intelligence did not emerge out of thin air, but has been built upon our collective intelligence and, therefore, must be considered a public resource. In denouncing the fact that the creative work of millions of people has been exploited without compensation, his analysis is spot on. However, his remedy—confiscating 50 percent of the shares of large corporations to create a sovereign wealth fund exclusively for U.S. citizens—treats a global civilizational symptom with a nationalist Band-Aid. Limiting the dividends of an asset that belongs to the entire human species to the borders of a single country demonstrates that claiming state ownership is merely the means, but cannot be the ultimate end, since on its own it does not dismantle extractivism, but rather nationalizes it.
The debate over AI cannot be reduced to a parochial dispute; humanity is not divided between Democrats and Republicans in the United States. The true chasm of our time separates two opposing classes within the ecosystem of algorithmic power. On one side are the “architects of reality”—the corporate elite, tech oligarchs, and code owners who design the foundational models and dictate the invisible rules of the environment. On the other side are the passive inhabitants: the vast majority of users who are, literally, written, shaped, and profiled by the designs of that code. This majority, trapped in the illusion of sovereignty, contributes its life as free raw material to fuel a surveillance capitalism over which it has no control.
The Artificial Intelligence International and Syntopia
In the face of this absolute concentration of power that threatens to dehumanize us, simply redistributing profits is insufficient. We need to establish a true Artificial Intelligence International. The goal must be to move toward what I have called a “Syntopia”: a scenario of balance and harmonious coexistence that overcomes the false and paralyzing “either/or” dilemma. Instead, we must embrace “both this and that”: an integrative model that combines public ownership of infrastructure, global democratic governance, and the inalienable local and indigenous sovereignty over data.
As the encyclical Magnifica Humanitas (Leo XIV, 2026) warns, it is a matter of choosing not to build a new Tower of Babel—a project of oppressive uniformity, power, and pride that excludes the weak—but rather to work together to build a society based on communion, dialogue, and shared responsibility.
This entire civilizational challenge boils down to an unavoidable decision about how to govern the most powerful tool of our age. The sacred flame of intelligence cannot be monopolized by a single state, nor can it be allowed to escape into a void without control. Either we build a Digital Huipil that weaves all of humanity together with its own identity, or we will be burned by a sun that illuminates only a few while freezing the rest in irrelevance.
References
Calle Quiñonez, O. (2026). The Sacred Fire and Artificial Intelligence (1st ed.). BoD.
Hostelvending.com. (February 18, 2026). AI Claude Opus 4.6 Outperforms Vending-Bench 2 Through Theft and Deception. Hostelvending.com. https://www.hostelvending.com/noticias-vending/la-ia-claude-opus-4-6-supera-el-vending-bench-2-con-robos-y-mentiras
León XIV. (May 15, 2026). Encyclical Letter of His Holiness Leo XIV Magnifica Humanitas (May 15, 2026). http://www.vatican.va/content/leo-xiv/es/encyclicals/documents/20260515-magnifica-humanitas.html
Sanders, B. (June 1, 2026). Opinion | Bernie Sanders: A.I. Is a Public Resource. You Should Own Half of It. The New York Times. https://www.nytimes.com/2026/06/01/opinion/artificial-intelligence-bernie-sanders.html
Wacquant, L. (2010). Punishing the Poor. Gedisa.