The Colonial Library Goes Digital: On Colonial Gnosis, Algorithms, and the Invention of the Digital “Other”

Private equity owns the cloud. But who owns the stories the cloud tells?

Hanna Barakat & Archival Images of AI + AIxDESIGN / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/

Private equity firms quietly control the infrastructures on which modern life depends. They fund a network of wealthy entrepreneurs and engineers whose intelligent machines are met with growing public unease.

Not coincidentally, this network is concentrated in the Global North, where billionaire developers race to make systems compute faster, simulate intimacy more convincingly, and scrape ever more “relevant” data. Shareholders play their paradoxical role as both master and jester, driven by the promise of portfolio expansion. Their only question to the technology company: How will you accelerate growth at the lowest possible cost and ensure your new system penetrates every industry vertical?

Somewhere along the way, these decisions produce what industry analysts call “unprecedented risks”, or worse, “unforeseen consequences.” But the damages are neither unprecedented nor unforeseen. To ordinary people, they were always fairly easy to anticipate.

“Ordinary” is itself a complex designation, shaped by the same currencies that bring one closer to, or further from, visibility. This is precisely where the problematization of technology transcends political alignment. A Republican voter may condemn Big Tech for mishandling online child grooming cases while simultaneously endorsing the tracking and abduction of immigrants.

If demoralization is a consequence of capitalism, we must also ask: what of the chip-factory workers in China forced to labor through lockdowns? What of the children in Katanga recruited to mine the cobalt that powers our devices? Technology companies occasionally acknowledge their oversights when harms come to light. They point to the imperfect “humanness” of technology development, hoping to better relate to the masses. But there has always been a blind spot in the history of tech.

That blind spot is colonialism.

Necropolitics and the Brutalist Mood

Critical inquiry into technology has long examined how convenience obscures the violent realities of production. For Achille Mbembe, this violence is not merely economic. It reflects what he calls the brutalist condition of our age: the erosion of boundaries between the genuine and the artificial, truth and non-truth, life and its technological double. Most alarmingly, it signals the dissolution of limits on violence itself.

Mbembe describes brutalism as a “form of power without external limits,” produced through the convergence of economic reason, digital reason, and biological reason. At the opposite pole lies what he terms collapsology: the slow combustion of the earth, a condition in which entire populations are forced to inhabit the ruins of worlds that have already collapsed.

Mapping these collapsing worlds requires moving beyond the simplicity of the North–South binary. Neocolonialism no longer operates solely through distant imperial centers. It is reproduced through local elites who internalize and replicate colonial logics in pursuit of legitimacy, capital, and recognition from what V.Y. Mudimbe calls “the father.” The colonial relation has not disappeared; it has been reorganized. Contemporary power increasingly operates through transnational circuits of capital, security, extraction, and technological governance that bind ruling classes across the Global North and South, while vast populations are rendered surplus to the demands of accumulation and control.

As Mbembe writes:

“The selection drive is now based on manifold forms of nanotechnology.”

By situating the hyper-digital future within these longer histories of domination, Mbembe reveals how technofetishism’s promises of superintelligence, engineered empathy, and algorithmic safety do not transcend colonialism. They often reproduce its underlying logics in new forms. This is why he argues for a new critical apparatus. Technology can no longer be understood as an external force intruding upon human life. It increasingly replicates, memorizes, archives, and downloads life itself, altering not only how we live but what life comes to mean.

The question, then, is no longer whether to advocate technology or oppose it. We already know that technological systems consume resources, reorder environments, and use bodies as fuel. The deeper question concerns the forms of life, memory, and desire they produce. In its aggressive back-and-forth with the human, networked technology ceases to function as a mere instrument. It becomes a repository of dreams, a keeper of memory, and increasingly, a trusted Other.

The Dream Machine

By recording our deepest curiosities and fears, AI becomes what Mbembe calls the “dream machine”, fed by our private data and stored in the “clouds.” These digital residues construct an unconscious that is reflected back on our screens, quietly steering our actions in the real world. We are paying a colossal price for this. The cost of externalizing our desires has metabolized into physical ruins: the targeted flattening of schools, libraries, markets, hospitals, homes, and olive trees in Occupied Palestine.

Surviving artwork at Al Qarara Cultural Museum, South-East of the Gaza Strip after it was bombed by Israel in Oct 2023 & Artwork Restoration attempts.

Manufacturing these ruins required turning violent dreams into economic reason, where fear becomes political data. Hyper-automated weapons that hunt “terrorists” borrow directly from colonial vocabularies that have always described certain people as “human animals” or “subhumans.” They belong to longer traditions of colonial dehumanization that make extraordinary violence appear both rational and necessary.

This logic is explicit. Donald Trump once bragged about arming Netanyahu for the genocide in Gaza:

“Bibi would call me so many times, can you get me this weapon, that weapon, some weapon I’d never heard of, and I made them. And [Netanyahu] used them well.”

US-manufactured Vision 60 robot dogs, deployed by Israeli military reserves in Gaza.

We don’t hate them enough. Not the robot dogs. The men who send them. The shareholders who profit. The empire that supplies the bombs and the politicians who pretend innocence. The unfolding of these desires widens the concept of necropolitics, the power to “make die.” But there is a deeper violence. Beyond stealing the right to live, neocolonialism has also stolen the right to dream.

The genocide in Gaza is part of a wider program of “making invisible” both life and imagined futures. It is connected to the deliberate destruction of libraries and cultural heritage sites, places where hopes are documented and exchanged. So we must disentangle the algorithmic rationalities fueling these weapons. Whose reasons were taken into account? Whose dreams? What gets erased, amplified, or censored when we download the world into a digital twin?

The Colonial Library

To understand how colonial logics persist within contemporary technology, we have to return to what V.Y. Mudimbe calls the colonial library. Colonial domination was never simply a matter of territorial conquest. It was also an organized project of translation, documentation, classification, and invention, a machinery of knowledge designed to render the world legible according to European categories.

At the heart of Mudimbe’s work lies a simple but devastating observation: the Other is always at risk of being mistranslated. He identifies three principal agents in this process, the explorer, the colonial administrator, and the anthropologist. Together, they participated in a vast project of normalization, reducing difference to a model established elsewhere.

This logic extended beyond politics and into science itself. Linnaean taxonomy classified Europeans as rational, inventive, and governed by law, while Africans were characterized as crafty, capricious, and governed by impulse. Colonial science implied that any lifeway outside Western institutions was savage or ignorant. Such classifications did not merely describe the world; they helped produce a hierarchy through which Europe could imagine itself as the measure of civilization. Diverse histories, cosmologies, and social worlds were translated into a language of lack.

As Zubairu Wai argues, Africa was constituted through the epistemic gaze of colonial modernity, a Europe that falsely projected its own cogito onto the African. In the name of civilization, enlightenment, and progress, complex societies were stripped of their histories and reintroduced to themselves through colonial categories.

The colonial library does not merely exclude. It invents. It transforms living worlds into objects of knowledge while eliminating what Mudimbe calls the “stressing elements”, the contradictions, meanings, memories, and forms of knowledge that resist incorporation into the colonial frame. We have never truly broken with this archive. The technologies may have changed, but the underlying operation remains strikingly familiar. What was once recorded in travel journals, missionary reports, censuses, and ethnographies is now encoded in datasets, classification systems, algorithms, and predictive models.

The colonial library did not disappear. It was digitized.

The Coded Gaze: Naming and Erasure

The colonial library persists not only through grand systems of classification but also through seemingly mundane acts of naming.

We see this today in the ways algorithmic systems interpret non-Western languages and identities. For example, Qassam, yes, an Arabic word associated with generosity, sharing, and division, was classified by Facebook as inappropriate due to the platform’s association of the term with the Al-Qassam Brigade. A birthday post celebrating an infant bearing the name was censored because the platform reduced the word to a single political association, severing it from the broader linguistic, cultural, and historical worlds from which it emerged.

Women named Isis encountered similar forms of erasure. Despite the name’s deep roots in ancient Egyptian history and mythology, social media platforms treated it primarily through the lens of contemporary security concerns, leading in some cases to account removals and restrictions.

These incidents are not mere technical errors. They reveal a recurring pattern. By stripping words of their histories, relationships, and meanings, algorithmic systems reproduce the same logic that animated the colonial library. Languages, symbols, and identities are detached from their contexts and reassembled according to categories established elsewhere. Complexity gives way to legibility. History is compressed into a glossary of fear.

This is precisely what Joy Buolamwini describes as the “coded gaze”: the values, assumptions, and exclusions embedded within technological systems that present themselves as neutral. Research by Timnit Gebru and Emily Bender further demonstrates that many harms associated with artificial intelligence originate not from machine autonomy but from structural inequalities embedded in training data. Large-scale datasets overwhelmingly reflect the perspectives most visible within digital infrastructures while excluding forms of knowledge that remain offline, marginalized, or underrepresented.

When intelligence is constructed from such uneven archives, the result is not universality. It is the projection of a particular worldview masquerading as universality. In this sense, AI inherits a familiar colonial impulse: the transformation of a partial perspective into a supposedly objective view of the world.

Toward a Decolonial Technology

Creating technology for all requires a rupture with capitalist accelerationism. It demands an ecosystem that genuinely treats technology as a collective human inheritance rather than a mechanism for extraction, surveillance, and accumulation.

Scholars such as Joy Buolamwini have described this struggle as a challenge to the “coded gaze”: the work of dismantling the electronic colonial library through participatory forms of research and technological development. Such efforts seek to build tools capable of engaging the languages, histories, philosophies, and lifeworlds that contemporary digital infrastructures routinely marginalize or exclude.

This work is urgent because decolonizing technology is not simply a technical task. It is an intellectual and political one. It requires confronting the colonial assumptions embedded within our systems of knowledge and the institutions that continue to reproduce them. Yet a more representative dataset alone will not solve the problem. The real challenge lies deeper.

Those who govern technological infrastructures must first reckon with their own colonial gnosis: the inherited assumptions, hierarchies, and ways of seeing that quietly shape what becomes intelligible, valuable, and actionable within digital systems.

The real work has little to do with data. It demands an ideological rupture. The data was there. The harms were foreseeable. The histories were available. And still, the colonial library persists, in server farms, predictive models, automated weapons, surveillance systems, and the silent censorship of a child’s name.

The question, then, is not whether artificial intelligence can become more intelligent.

It is whether those who occupy positions of epistemic, economic, and technological power can cease projecting their own image onto the rest of humanity.

Can those in the occupying spaces of power finally abandon the colonial impulse to invent the Other? Or will the cloud continue to remember only the dreams of empire?