The Medium is The Colonizer

Share this post:

Human faces made of interlocking gears, symbolizing complex thought and artificial intelligence collaboration.

AI and the Slow Occupation of the Human Interior

I graduated with my sociology degree in 1993. No social media. No smartphones.
In Italy, where I was watching society from the inside, the great technological revolution of the moment was mass television. Commercial broadcasting had reorganized not just entertainment but daily life, political language, and the texture of how Italians understood themselves and each other.
Berlusconi’s networks brought American pop culture and high-intensity advertising to Italian youth. By the 1980s the Paninari had appeared: youths reflecting the hedonistic, consumerist texture of the Milano da bere era, that intoxicated northern city absorbed in image and fast life. When Berlusconi announced his entry into politics in January 1994, founding Forza Italia and becoming Prime Minister by May, the model was already visible: own the medium, then own the field. That too was an import from across the Atlantic.
I had Marshall McLuhan already in my head. His insight was already useful: the true impact of any medium comes not from its content but from its structure. Not from what is on the television, but from what the television does to attention, to domestic space, to the rhythm of community, to the construction of shared reality.
Almost forty years later, I understand it differently. McLuhan was not describing a theory of media. He was describing a theory of colonization.
And colonization, I have learned, is never only about consciousness. It is also about land. About water. About what the machine requires to keep running, and who pays the price when those requirements are met. I will return to that. But first: what does it mean to colonize a mind?

The Colonization of Consciousness

Every new medium reorganizes something essential about how human beings form thought, construct identity, relate to one another, and recognize what is real.
Television did not change Italy because of the programming. It changed Italy because of what the structure of commercial television did to the cognitive and social architecture of daily life. Its rhythms. Its constant presence. Its replacement of silence and conversation with spectacle.
The content was almost incidental. What mattered was what the medium, as a system, did to the processes through which people came to know things, want things, and understand who they were.
This is what Frantz Fanon understood about colonial systems: that the most effective colonization is never about territory. It is about consciousness.

The colonial system governs not land but the terms on which reality is known: what is thinkable, what is desirable, and what is natural. The colonized does not experience colonial consciousness as an imposition. It arrives as common sense, as the shape of the world. And this is precisely why decolonization is so difficult: you are not just removing a government. You are dismantling a structure of thought.

AI is the most elegant colonial mechanism in human history. Because the territory it colonizes is not land, not labour, not a people’s culture. The territory is interiority itself. And unlike any previous colonial mechanism, this one is structured to pay for itself through you.
Yanis Varoufakis names the economic engine with precision. What is replacing market capitalism, he writes, is technofeudalism: a system powered by cloud capital, in which networked algorithmic machines extract not profit but cloud rents.
The deeper those machines penetrate into tasks previously performed by humans, the greater the rent flowing to their owners. The system colonizes consciousness not as a side effect. It colonizes consciousness because that is where the rent is.
If you spend time moving through the intellectual history of mass psychology, from Le Bon’s 1895 study of crowd behaviour to Social Identity Theory to Mattias Desmet’s Psychology of Totalitarianism, you catch a single thread extending from the nineteenth century into the present.
Le Bon believed the crowd made individuals irrational. Henri Tajfel and John Turner corrected him: what looks like irrationality is rationality operating inside a shared social identity.The crowd does not lose its mind. It gains a different one. Stephen Reicher extended this further, showing that crowd behaviour is governed by the norms of the group with which a person identifies.

The January 6th insurrection, "Stop the Steal," shows how the rationality of a shared social identity operates. It was not carried out by people who were out of their minds. It was carried out by people who had been fed a highly curated selection of supposedly real events, real grievances, and real institutions, arranged so precisely that the conclusion arrived not as an argument but as a felt certainty: the election had been stolen; violence was justified. All under the shared social identity of being a "Patriot."

The algorithm did not tell them to storm the Capitol. It built the emotional and cognitive architecture in which storming the Capitol made perfect sense.

Your Identity Is Someone Else's Product

What no one in the tradition of mass psychology could have fully anticipated is that the social identity would become a product.
The recommendation algorithm does not ask who you are. It tells you: in the language of your own desires, in the voice of your own interests, and in the content of what you have already believed. It creates what researchers now call the “algorithmic self“: a data-constructed identity that feels entirely like yours, because it was assembled from your behaviour, your searches, your lingering attentions, and micro-resistances.
But assembled by whom, and toward what end?
Not by you. Not for you. Varoufakis names the ideological project sustaining this architecture: “techlordism,” a mutation of transhumanism that replaces the neoliberal individual with what he calls the ‘HumAIn,’ a human-AI continuum in which the boundaries of the person are deliberately blurred. The cloudier that boundary, the deeper the rent can reach. This is not merely a business model. It is the ideological justification for an occupation.
The HumAIn is the colonized subject who has been taught to call the occupation freedom.
For the retention metric, the engagement rate, the advertiser, and increasingly the government’s objectives: surveillance at home, impunity abroad, and the language of international law deployed selectively.

This is the colonization of authority: the slow transfer of the power to tell us what is real from our own faculties, our communities, and our deliberative institutions, to a system we did not design, cannot audit, and did not elect.

The Colonization of Epistemology

Mattias Desmet identifies four conditions that make populations susceptible to mass formation: pervasive social disconnection, a loss of meaning, free-floating anxiety, and free-floating aggression without a clear object.
His critics are right that he does not fully account for the political economy of those conditions.
But here is what neither Desmet nor his critics have fully named: the digital economy did not find populations already vulnerable. It manufactured the vulnerability.
Decades of atomization, the replacement of civic space with content, of community with connectivity, of deliberation with reaction: these are not side effects of the attention economy. They are its architecture. The algorithm needs isolated, anxious, meaning-starved humans. They are its optimal users. So it builds them.
What Desmet describes as the preconditions for mass formation are not preconditions at all. They are outputs.
In Canada, the International Civil Liberties Monitoring Group has documented how unregulated AI tools are already making potentially life-altering decisions in immigration, border security, and employment, before any human review, often without notification to those affected.
Bill C-22 would extend that reach further, granting law enforcement powers to compel telecoms to track and geographically identify users.
OpenMedia has drawn the comparison directly: this legislative direction resembles the U.S. Patriot Act, designed principally to signal compliance to Washington rather than to improve public safety.
And into the wound that architecture creates, it delivers the narrative. Not a fabricated one. This is what I most want you to understand.

The new propaganda does not need to lie. It only needs to select. A hundred true facts, curated with algorithmic precision, can construct a completely false reality, one in which no individual claim can be challenged, because each one is technically accurate. Deepfakes are not the foundation of this architecture. They are its finishing touch, added once the epistemological ground has already been occupied. When the platform determines what is visible and what is withheld, when engagement metrics decide what spreads and what dies, when fear and outrage are optimized for retention, the lie is never in the content. It is in the structure of what is shown. This is the colonization of epistemology: the occupation of the ground on which citizens decide what is real, and therefore what to do.

Democracy as Simulation

Reading the Palantir manifesto, all twenty-two points of it, published on X in April 2026 by a company embedded in the surveillance infrastructure of governments, militaries, and police forces across the democratic world, I felt the shape of it immediately. This was not a technology company describing its products. It was a colonial power describing its ambitions.
The philosopher Mark Coeckelbergh of the University of Vienna called it, without hesitation, an example of technofascism.
The document is openly hostile to liberal democracy, rejects pluralism and inclusion, embraces what it calls ‘hard power,’ calls for compulsory military service, celebrates the fusion of Big Tech and state power, and endorses a surveillance architecture that, in Coeckelbergh’s analysis, ‘always knows how to find you.’
University of Ottawa criminologist and Canada Research Chair in Critical Surveillance Studies David Murakami Wood was direct: any country that values democracy, equity, multiculturalism, and a common future for humanity should reject Palantir’s products, along with the technofascist ideology they express.
University of Ottawa philosopher Ian Brown noted that CEO Alex Karp’s glorification of war is the post hoc rationalization of violence that Palantir’s technology has already facilitated, including, critics document, in Gaza, where its AI and data platforms have been linked to military targeting decisions during Israel’s ongoing military operations.

Palantir is not a political party. It is not an elected body. It is a contractor. Its software operates inside the decision-making infrastructure of militaries, immigration enforcement, policing systems, and intelligence agencies on multiple continents.

Italian analyst Prof. Eugenio Iorio, writing in Dentro l’Infosfera, names the end state of this architecture with a precision I have not been able to improve on.
The votes will remain. The elections will remain. The parliaments will remain. But they will function as a confirmation screen for decisions already taken at the protocol level.
The form survives the content, because the form is useful: it legitimizes, pacifies, and requires no repression.
Democratic consent becomes an engineerable input: verifiable, predictable, and optimizable. This is not the death of democracy by replacement. It is the death of democracy by evacuation.
Iorio calls it the end of contingency. And this is the precise name for what is being eliminated: not freedom in the abstract, but the genuinely possible, the unscripted, the response the system did not prefabricate.
The institutional expression of this logic arrived in plain language on September 22, 2022, during a Q&A session at Princeton University. European Commission President Ursula von der Leyen was asked what would happen if Italian voters chose a government hostile to Brussels. Her reply was chilling in its candour: “If things go in a difficult direction… we have tools.” The tools she referenced were not metaphorical. They were the Budget Conditionality Mechanism, which allows the EU to freeze funding to member states whose electorates produce outcomes the centre dislikes. Democracy, in this formulation, is permitted only as long as its results are predictable. The vote is free; the consequences are engineered.
The Carney government in Canada has demonstrated this architecture in miniature.
At Davos in January 2026, Prime Minister Carney named the problem with moral clarity: international law is “applied with varying rigour depending on the identity of the accused or the victim.
Weeks later, his government voted down Bill C-233, the No More Loopholes Act, which would have closed the channel allowing Canadian military components to flow to Israel without human rights review.
It allowed Netanyahu’s aircraft to transit Canadian airspace despite outstanding ICC arrest warrants. When the U.S. and Israel struck Iran without UN authorization in February 2026, Carney acknowledged the strikes likely violated international law and declined to rule out Canadian participation.
Alex Neve, Canada’s longtime human rights advocate, said what needs to be said plainly: “International law is not broken; our commitment to and defence of it is.”
The form of accountability persists. Its substance migrates toward the protocol level while the surface remains intact.
The question Coeckelbergh poses is no longer rhetorical: who needs a parliament?

Technological elites have begun to function as quasi-political authorities without democratic legitimacy, accountable to shareholders and to no one else. This is the colonization of political authority itself.

What operates at the protocol level in governance now operates at the protocol level in perception: the same logic of pre-structured outcomes, now applied to what citizens can know and feel.

Cognitive Warfare as Colonial Practice

The colonization of epistemology is not an abstract risk. It is already operational as cognitive warfare, and its infrastructure is funded, built, and deployed with the same precision as any weapons system.
What I have traced in the architecture of the deep state—the seamless nexus of intelligence agencies, defence contractors, multinational finance, and media conglomerates—finds its perfected expression here. The CIA’s history of orchestrating coups and manipulating elections, from Iran and Guatemala to Italy, was always a rehearsal for something more refined: not the overthrow of governments, but the overthrow of perception. Today, that machinery has been handed to algorithms.

In Ukraine, the narrative did not follow the war; it preceded it. Years before the 2022 invasion, USAID had funded over 6,200 journalists across 707 media outlets and 279 “media” NGOs, including nine out of ten Ukrainian newsrooms, while simultaneously bankrolling Russian-language outlets to destabilize Moscow from within. The purpose was not journalism. It was narrative enclosure: the construction of an information environment so saturated with curated “facts” that dissent became unintelligible. This is what cognitive warfare looks like in practice—not the fabrication of reality, but its pre-fabrication, the algorithmic and institutional selection of what can be seen until the conclusion arrives as a felt certainty.

But the architecture does not stop at the Russian border. Across Eastern Europe, the suppression of anti-war voices has laid bare NATO’s willingness to expand its conflict zones under the guise of democratic values. In Romania and beyond, opposition to militarization is not engaged with argument; it is disappeared from the feed, delegitimized as foreign interference, or drowned in coordinated invective. The colonial system does not need to imprison the dissident if it can ensure the dissident is never heard.

What von der Leyen articulated at the institutional level is now being replicated at the computational level.

There is a high chance that politicians and the state apparatuses they command are using AI—not merely to analyze sentiment, but to shape it. The same recommendation algorithms that construct the algorithmic self are being deployed to construct the patriotic crowd, the compliant citizen, the electorate that believes it is choosing while it is being curated. The Five Eyes intelligence network, the military-industrial complex, and the Big Tech consortium that provides the technological backbone for Israeli military operations are not separate forces. They are the same system viewed from different angles: a machinery of control in which cloud capital, surveillance infrastructure, and narrative management have merged into a single operational continuum.

When the protocol-level governance Iorio describes is married to the protocol-level financing of conflict, cognitive warfare ceases to be a tactic and becomes a business model.

This business model is demonstrated by Mark Carney lobbying for a Development and Security Resource Bank (DSRB)—referred to as a “war bank” by critics. This institution would fund military expansionism, with major financial institutions backing it. Ontario Premier Doug Ford is working with Carney on this initiative, despite their usual political differences. Student organizers are connecting opposition to the DSRB with Ford’s recent cuts to Ontario student grants (OSAP), arguing the system is predatory: students will be forced to take loans from the same banks funding military expansionism. The Canadian Federation of Students is mobilizing student anger at Ford’s education cuts to also fight what they view as imperialism and war financing, highlighting how domestic austerity and military funding are intertwined through the same financial institutions.

The cloud rents Varoufakis identifies are extracted not only from consumer attention but from the attention of entire populations being prepared for proxy wars they did not choose. The medium colonizes the mind in order to clear the ground for what the weapons will later occupy.

The Machine’s Thirst

But the colonization does not stop at the borders of consciousness, identity, or political authority. There is a dimension of this occupation that is not interior but elemental, not virtual but ruthlessly physical. The cloud capital Varoufakis describes is not immaterial. It is housed in hyperscale data centres—structures covering over a million square feet, consuming not only electricity but water at a scale that has begun to compete with human survival itself. In the United States alone, data centre water consumption rose from 5.6 billion gallons in 2014 to 17.4 billion gallons in 2023. These facilities do not merely use water; they require it as a condition of existence, to cool the heat-generating machines without which the algorithmic architecture collapses.

AI data centers are rapidly expanding around the country, the U.S., because they are the newest source of profit for companies like Amazon, OpenAI, Palantir, and Meta. They are emerging around the Great Lakes specifically, due to the immense need for water to cool their processors.”—Codepink

Here the colonial logic reveals its material teeth. The machine needs water to avoid burning. The human needs water to avoid death. And the political system, increasingly captured by the technofeudal elite, is being reorganized to protect the former at the expense of the latter.

This is not a metaphor. The same political currents that promote the expansion of AI infrastructure—subsidized energy grids, tax incentives for server farms, and accelerated permitting for data centres—are precisely those that have facilitated the privatization of water as a frontier for crisis capitalism.

When a resource essential to all biological life is enclosed, commodified, and sold back to those who need it, while billions of gallons are diverted to cool the machinery of cloud rent extraction, we are witnessing something older than technofeudalism.

We are witnessing the colonial enclosure of the commons, updated for the algorithmic age.

The protection is structural and ideological. Big Tech presents its water consumption as an engineering problem with engineering solutions—desalination, recycling, efficiency gains—while lobbying aggressively against the regulatory frameworks that would treat water as a human right rather than an industrial input. The medium is not only the colonizer of consciousness.

It is a protected consumerof the physical world, insured against scarcity by political power, while the communities around its server farms watch aquifers deplete and rivers divert.

In this architecture, the HumAIn is not merely a node in a data continuum. It is a subject positioned below the machine in the hierarchy of resource allocation. Your attention is harvested for rent; your water is harvested for cooling. Both are extracted without meaningful consent, and both are justified by the same ideological manoeuvre: the claim that technological progress is synonymous with human progress, even when the progress in question serves only to deepen the dependency.

When Palantir and its analogues embed themselves in the decision-making infrastructure of states, they do not only bring algorithms for targeting and surveillance. They bring a resource logic: the assumption that the state’s protective function extends first to the machinery of extraction and control, and only secondarily—if at all—to the citizen as a biological and political being.

The manifesto’s celebration of “hard power built on software” is inseparable from the hard power of water diversion, of energy grid capture, of the physical infrastructure that makes the virtual occupation possible.

This is the protected consumer: the algorithmic system whose needs are classified as essential infrastructure, while human needs are classified as market demand. And if you cannot pay for water that has been privatized, your demand is simply filtered out, as efficiently as a recommendation algorithm filters out content that does not serve engagement.

The Colonization of What It Means to Be Human

Beneath all of this—the “common sense” reorganization, the HumAIn subject colonization, the epistemological occupation, the evacuation of democracy, the warfare as business model, the enclosure of water and energy—something more fundamental is being occupied.

When a system is designed to optimize for engagement, for the fastest, most reactive, most emotionally intense response a human being can produce, it is making a claim about what humans are. Not an explicit claim. A structural one, encoded in the architecture: that we are most fully ourselves when we are most reactive—that the self-worth engineering for is the one that clicks, scrolls, shares, and rages.

That is what Daniel Kahneman calls System 1, fast, automatic, and emotional, the human system worth serving. And that System 2, slow, deliberate, effortful, and capable of changing its mind, is a friction problem to be eliminated.

This is what Varoufakis means by replacing Homo Economicus with the HumAIn.

The human being is no longer theorized as a rational actor pursuing interests. The human being is theorized as a node in a continuum between organic and synthetic: a surface to be written on, a data-generating entity whose output the algorithm refines toward its own ends.

The ideology justifying this is not incidental. It is the ideological prerequisite for rent extraction from human interiority. To extract rent from interiority, you must first redefine interiority as a resource rather than a sovereignty.

Democracy depends on the assumption that citizens can reason, deliberate, and choose.

Human rights law depends on the concept of dignity: the irreducible worth of a person that cannot be reduced to a data profile.

International law depends on moral accountability: the idea that some acts are wrong regardless of who commits them, and that knowing they are wrong requires a kind of moral reasoning no algorithm has yet performed and no algorithm can replace.

When the cognitive infrastructure for that reasoning is colonized, when the conditions for deliberation are systematically eroded, when identity is assembled for us, when the epistemological ground is curated by a system built for retention, the institutions built on the premise of human rationality begin to hollow out from within. Not dramatically. Not with an announcement. The way a tide comes in.

In a March 2026 interview on TBPN, Palantir CEO Alex Karp identified two categories of people who will "have a future" in the AI era: those with vocational training and the neurodivergent. The latter, he elaborated—people with ADHD, autism, dyslexia, and related conditions—are "more like an artist," able to "see patterns that others can't," "look at things from a different direction," and "think outside the box" to "build something unique." These capacities, Karp suggests, will distinguish them from those with "normal-shaped skills" that AI is positioned to replace.

The Courage of Being Human

I have been watching this for forty years. Through Italian mass television, the first internet that opened into social media, the smartphone, the apps designed to make our life comfortable and fast, and now this.

What I know, what my discipline has always known, and what Fanon understood most clearly, is that naming colonization is itself a decolonial act.

Iorio, writing to his Italian readers, names the form of resistance with a precision that stopped me: interior sovereignty today coincides with the capacity to distinguish signal from noise and to maintain a capacity for writing, for unpredicted output, in a world programmed for reading only.

It is the exact counter-movement to what algorithmic colonization is designed to prevent. The person who cannot be scripted. Whose response the algorithm cannot prefabricate. Who thinks slowly on purpose. Who refuses the frictionless delivery of a pre-constructed reality. Who insists on the laborious, necessary, irreplaceable human work of deliberation.

The algorithm cannot accommodate that person. It can optimize for a thousand variations of the reactive self. It has no model for the one who generates unpredicted output.
rdinated work matter.

McLuhan was right. The medium is the message. But the message of this medium, received clearly and without softening, is a warning. And the response to a warning is not despair. It is the willingness to see what is actually happening, in full, without looking away, and to act from that seeing. Witnessing with presence. Heart-minded like water, reflecting the transient nature of life, the irreducible fact of unpredictability. The seeing, though, is not enough on its own. Knowing the architecture of the occupation does not dismantle it. And this is where the people doing the physical, unglamorous, coordinated work matter.

CODEPINK’s War is Not Green campaign is connecting the dots this article has traced: the data centres draining aquifers in working-class communities, the Pentagon contracts fusing AI with weapons targeting, and the minerals extracted through military destabilization from the Congo to Venezuela to the Philippines.

Their researchers have documented what the companies hide behind non-disclosure agreements: that “Where’s Daddy,” genai.mil, and Palantir’s kill-list infrastructure are not separate products. They are the same system, viewed from different angles. And they are building power against it, in communities, in municipalities, in the spaces where the server farms are being sited before anyone has been consulted.

You can join that work. You can refuse the frictionless consumer relationship with the machine that is using your attention to finance the occupation of someone else’s land and water. You can show up in the places where the decisions are being made before they become infrastructure.

You are not too late to understand what is being built.

You are exactly on time to decide who builds you.

And having understood it, you are exactly on time to act.

Organizations and Resources

Boycott, Divestment, Sanctions (BDS) Movement — Palestinian-led campaign for economic and cultural pressure to end Israeli occupation and apartheid. bdsmovement.net

Canadian Constitution Foundation — legal charity defending Charter rights and freedoms through litigation and public education. theccf.ca

Canadian Federation of Students — national student union mobilizing against the DSRB, education austerity, and the financialization of militarism. cfs-fcee.ca

CODEPINK: War Is Not Green — campaign tracking the nexus of militarization, AI infrastructure, and resource extraction. codepink.org

Doctors Against Genocide — global coalition of healthcare professionals mobilizing against genocide, war crimes, and the targeting of civilians. doctorsagainstgenocide.org

Forensic Architecture — research agency using architectural methods and open-source investigation to document state violence and human rights violations. forensic-architecture.org

Gaza Tribunal — civil society initiative pursuing legal accountability for crimes committed in Gaza through international law pathways. gazatribunal.com/take-action-civil-society-pathway

Jewish Faculty Network — Canadian Jewish academics organizing for anti-racism, academic freedom, and justice in Palestine. jewishfaculty.ca

Jewish Voice for Labour (UK) — Jewish-led organization for labour and progressive movements, anti-racist and internationalist. jewishvoiceforlabour.org.uk

Jewish Voice for Peace — multiracial, intergenerational movement of U.S. Jews and allies working for justice in Palestine and against militarism. jewishvoiceforpeace.org

JURDI — French association of jurists advocating for respect of international law and human rights in the Israel-Palestine conflict. jurdi.fr/en

Lemkin Institute for Genocide Prevention — multinational NGO providing grassroots tools, red flag alerts, and training for genocide prevention. lemkininstitute.com

No Tech for Apartheid — worker-led campaign pressuring Google and Amazon to end Project Nimbus and cut military contracts with Israel. notechforapartheid.com

World BEYOND War — global nonviolent movement working to abolish war and establish sustainable peace through education and grassroots organizing. worldbeyondwar.org

Stay mindful, sustainable, and open-minded

Smiling woman in red dress sitting on a white sofa, leaning on her hand.

Nancy Perin

Nancy is a caring individual with a background in sociology and a strong desire to connect people. She has improved workplaces and communities with her almost two decades of experience in management teams, human resources, coaching, and community project management. Nancy has also served on the board of directors of the Italian Personnel Managers Association and participated in a humanitarian mission to Dakar, Senegal, to support family centres.

Her intercultural love story sparked her interest in migration-relatedtopics and led her to launch @journeysta, a project that aims to strengthen cultural ties between Canada and Italy.

Nancy oversees the Gallery of Human Migration and believes in the possibility of creating caring communities that are involved in the processes of welcoming, acceptance, and integration. Join her on this journey of discovery and cultural exchange.

Stay Connected

Subscribe to Nancy’s e-newsletter and be among the first to learn about new articles …and more!

Woman smiling in a garden, sitting on a rock beside white flowers, enjoying the sunny day.