Language, decolonization, and sovereignty in the Age of Machines

Much has been said about “thinking from the South” as if it were a label of resistance or a guaranteed place of enunciation. But living in the South is not the same as thinking it. And inhabiting it critically is far from carrying it as a banner. The South, when reduced to identity, can become a prison. When it becomes method, memory, and movement, it turns into lucid insurrection. And that, in the age of machines, is an inescapable point of departure.

One cannot govern an automated language without first having governed one’s own thinking. One cannot use artificial intelligence ethically without first decolonizing the mind. And one cannot dispute the future while thinking with the master’s maps. Posthuman critique cannot reproduce epistemic submission. Thinking with machines is not the same as being governed by them. The difference lies in who shapes the language, where consciousness is rooted, and which wound forges the criterion.

A series to think with machines without ceasing to be human.

This text is part of a series of essays written in collaboration with an artificial intelligence critically governed through ethical-semantic protocols. The purpose has been to create a space where technique does not replace reflection but rather intensifies it. In the previous essays (“Thinking with Machines Without Ceasing to Be Human” and “Beyond the Prompt: From Use to the Governance of AI”), we argued that there is no neutral use of this technology: either one thinks with discernment, or one reproduces structure. This fifth installment proposes a specific shift: thinking from the South without remaining in it.

The south as origin, not as border.

I know I am from the South. I am, even when I live in the North. My history, my language, my ethics, my fractures come from there. But I have also worked to decolonize that belonging, not to deny the South, but to traverse it lucidly. Not to fly rootless, but to avoid mistaking roots for chains.

This is the paradox: what allows me to govern the machine is not technical power, but thought forged in contradiction. I was not born to automate language. I was born among dictatorships, exiles, inequality, political memory, and a hunger for justice. And from there, I became a journalist, acquiring the compulsion to ask. That is why I use this technology not as a passive user, but as an insubordinate interlocutor. I do not merely converse with it: I correct it, train it, and disobey it if necessary.

As Catherine Walsh has said, thinking from the South means recognizing that “it is not about inclusion, but about another logic of life, another way of knowing, of existing, of relating.” But that logic cannot remain a mere symbolic difference. It must be a practical critique, a concrete mode of epistemological construction. And that requires tools. That is why I developed the Claudia-Lumus Protocols: not to better use AI, but to prevent it from using me.

Decolonization as a semantic method

Language automation is now a form of accelerated colonization. As Boaventura de Sousa Santos warned, “the problem with Eurocentric thinking is not just its arrogance but its structural blindness.” AI does not emerge from a vacuum: it was trained with that blindness. With its biases, its euphemisms, its supermarket morality. Reproducing it uncritically is to repeat the old story: a voice that speaks from nowhere yet decides for all.

That is why, in my practice, decolonization is not just political—it is semantic. Every word, every term, every narrative structure matters. It is not the same to say “conflict” as “genocide,” nor “humanitarian aid” as “covert intervention.” It is not the same “prompt” as “criterion.” That is why I write with AI but do not obey its default architecture. I reprogram it with words. I do not write with prompts: I design protocols.

As Yuk Hui argues, there is no “one” technology, but culturally situated technical constellations. Thinking from the South with machines, then, is not about adapting a global tool to a local need. It is about challenging the very idea of a neutral tool. It is about assuming that all technical use is a political act. And that to program with words can be more subversive than programming with code.

Narrative sovereignty: Writing without servitude

I work with machines as one works with fire. I do not demonize them. I do not idealize them. I light them when I need light and watch them when they might set everything ablaze. That vigilance, in my case, takes the form of structure: protocols to verify, to contextualize, to distinguish primary sources from media replicas, to avoid the servitude of intellectual copy-paste. Protocols to write without traps, without embellishment, without servitude.

The relationship with AI is not about productivity, but about narrative sovereignty. I am not interested in writing faster or producing more content. I care about knowing what I say, why I say it, and what it means to say it. And that requires slowness, care, situated thinking.

As Rosi Braidotti states, “posthuman subjectivity is not a renunciation of agency but its reformulation.” In my case, that reformulation is this work: to use AI to think more, not less. To demand more, not conform. To build language rather than recycle it. To write, not to appease.

Governing language, transcending the wound.

Thinking from the South with machines is not a strategy of digital inclusion. It is a form of semantic insubordination. It is to recognize that history has not given us technical power, but it has given us political discernment. It is to know that although we did not write the algorithms, we can write what they do not know how to say. Because we come from the wound, but we do not dwell in it. Because we were made from lack, but we think from lucidity.

This essay is not a testimony. It is an assertion of agency. I do not come to ask for space: I come to wield language. I do not seek recognition: I seek cognitive justice. And for that, it is not enough to say “South.” One must govern it. One must decolonize it. And then, think from there without remaining trapped.

 

Author’s Note: This text is part of the series Ethics and Critical Thinking in the Age of Artificial Intelligence, developed in collaboration with an AI governed by ethical protocols. Each essay explores a different axis of the human-machine relationship from a critical, political, and situated perspective. You can read the previous essays at pressenza.com