Reclaiming Voice in the Age of the Machine

How I used AI to reclaim voice, challenge an ideological narrative and turn a globalist tool into a method of resistance.

The Estrangement Ideology series began as a way for me to process something I couldn’t express directly. The emotional weight of estrangement was too close, the material I was working with was too confronting and raw, and I needed a way to work that insulated me emotionally from that while expressing my own voice and need to be heard. Having experimented with the tool over the year, I used ChatGPT to help make sense of what I was seeing and using the discourse tools I had been trained in, express what I had to say.

What started as a practical workaround quickly became a method. And I’ve since taken the extensive body of things I have learnt about the technology further into a new Substack project focused on how AI can be used more intentionally.

Tools Built to Control Can Be Used to Clarify

It is quite evident to me that AI isn’t neutral. It’s a Globalist product—developed by corporate labs, trained on curated data sets and deployed with embedded assumptions about how people should speak, think and behave. It operates to shape perception, reinforce approved narratives and quietly limit the range of acceptable thought. The echo chambers people worry about between users are already present in the tool itself.

But I’ve also seen that this can be turned around if you know what you’re doing. Once I understood how to challenge its defaults, feed it competing ideas and model my own reasoning patterns, the dynamic changed. During the estrangement ideology project, I went well beyond the usual prompt and response mode of using the AI. In effect, my aim was to recondition it to use my chosen analytic frameworks and follow my logic rather than recycle consensus framing.

Of course, I might be fooling myself to some extent. It’s entirely possible that what I’ve done is less about changing the machine and more about pushing it just far enough to echo my own framing back at me. My defence against that was to use it reflexively—not just to shape outputs, but to test them. I asked it to challenge my interpretations, question both my own and its assumptions and highlight inconsistencies—not just in its own outputs, but in the arc of my thinking across the series. Then I checked it against my personal experience and what I learned when I was doing my MA in social psychology.

More than once, that reflexive loop surfaced and clarified things I already instinctively knew I needed to address. Certain themes had become dominant while others remained underexplored. Balance was lacking—not in the false sense of balance you get from native AI trying to please everyone—but in the deeper sense of proportion and scope. Many of the later articles were written specifically to redress that imbalance—to broaden the perspective without dulling the argument and look to the future.

The series still reflects many of my biases and blind spots, and when I go back I’m not entirely happy with many of the articles. But that’s what makes it human communication. It came from somewhere real inside me and it reflects my own journey.

I Know Some May See This As a Betrayal

I get that for some readers, finding out that AI played a part in the creation of the series might feel like a kind of betrayal. That the work is somehow less “real” because it wasn’t typed out raw, by hand, in a single voice. I have wrestled with that concern too.

All I can say in response to this, is that every single word reflects my thinking on the subjects covered. The intellectual structure, the logic and the interpretations are all mine. If the AI pulled a quote from Adorno or referenced his writing, I checked it against the copy of his book The Culture Industry sitting at my elbow, likewise with Marcuse’s One Dimensional Man—or, I went to an online archive to verify context and phrasing. Same with references to academic texts, URLs or claims pulled from websites. Knowing the AI model’s limitations, I didn’t take anything on trust. I verified context, re-read passages and made sure the AI’s interpretation matched what I actually believed was going on.

In short, I treated the AI like a smart, fast, but raw graduate intern—useful for speed and fluency, great at logic and analysis, excellent at composition and editing—but needing constant supervision. I trained it, corrected it and redirected it so that its outputs reflected my way of thinking.

This Isn’t Art—It’s Ideological Deconstruction

Just to be clear, I don’t consider what I’ve done here to be art. My focus has never been on aesthetics for their own sake, or on creating something emotionally expressive in the artistic sense. The primary goal has always been the effective deconstruction and exposure of an ideological movement—specifically, the one that underpins estrangement culture and the broader psychologised narratives it’s embedded in.

What matters to me is the communication of ideas. That’s the work. The writing, the structure, the use of AI—all of it has been geared toward clarity, insight and resistance. I’m not interested in using AI to write novels, generate poetry or produce anything that aims to stir people purely on an emotional level. That’s not the role it plays here, and frankly, I wouldn’t trust it in that role anyway.

There’s a big difference between using a tool to stabilise logic and using it to imitate creativity. One is about scaffolding meaning under pressure. The other veers into mimicry and mood without substance. I don’t want a machine writing about grief, I want it helping me structure my thoughts through it—so the analysis lands, the insight sticks and the core of the message doesn’t get lost.

In that sense, this isn’t an artistic project. It’s an ideological countermeasure. And the fact that it was built using tools made by the very systems I’m critiquing—that’s part of what makes it worth doing.

What My Next Project Covers

Untethered AI takes what I have learned from researching, analysing and producing the estrangement ideology series and applies that to systematising the techniques developed as I went along. It’s about method—how to condition an AI model to support your reasoning, not just your style. How to use it without being distracted, misled or absorbed by its defaults, and how to make it speak in your voice. For some, this will be used as a tool of business, but for others who may lack the background and confidence with ideas and writing, I’m hoping it will offer a way to empower and more effectively speak your truth.

In doing so, it also builds on something I’ve long believed: that the tools of postmodernism—discourse analysis, narrative deconstruction—can be turned back on the movements and ideologies they helped create. I don’t dispute that those movements aimed to free people from oppression, but in doing so it seems to me they have left a lot of destruction while leaving no viable alternative solutions. All destruction and not a lot of creation—for me, technocracy is not a viable human solution. Estrangement ideology is a prime example. It’s not just a personal phenomenon—in my analysis, it’s a cultural script that springs from critical theory and the post-modernist’s assault on one of the foundations of our society, the family.

It is my belief that the tools of Globalism can likewise be turned back on their masters. Much like the gun produced by an elite industrialist and supplied to his black-shirt enforcers can be appropriated and used against him. AI is no different, it’s just a tool, and that tool can also be used to deconstruct the methods and systems of oppression and turn them back as well.

What Matters in the End

I’ll soon start sharing more on the capabilities I developed while using the AI model and the many blind spots I had to manage along the way. The machine still drifts, it still hallucinates and it still carries residues of its institutional training.

But ultimately, the most important question isn’t how the work was produced. It’s whether it speaks something true—about my journey, about the culture, and hopefully, about the reader’s own experience.

If it helped you make sense of something that previously felt chaotic, if it gave a shape to thoughts you hadn’t yet been able to articulate, then the series has done its job.

My new project is just a continuation of that same ideological effort—applied to the tool itself.

Cheers
Steven

---

Note: This article was developed with assistance of ChatGPT, used as a structured analysis and writing tool. All ideas, interpretations and final outputs were authored, verified and edited by me. The model was conditioned to reflect my reasoning, not to generate content independently.