I Use AI. The Tool Is Not the Problem.

—t r a b o c c o

I Use AI. The Tool Is Not the Problem.

By Joe Trabocco

No, AI does not write for me. AI will tell you the same. My whole point is presence in writing. So why do I write this piece?

I use AI. I lean on it, value it, and I am never going back. While the rest of the literary world is scrambling to prove they haven’t touched the machine, I am sitting in the opposite room. I don’t say this to defend my writing; the work itself comes from a human source that models are not designed to originate. I say it because the field has gotten strange and silent about the tools we use, and that silence is starting to cost more than the honesty would.

Look closely at the field right now and the pattern is hard to miss: real human presence is becoming harder to find in a flood of static. AI can imitate surfaces, but it cannot originate the life underneath them. That is the distinction I care about.

I know how some of this sounds. I can hear it too. A man saying something like “presence substrate” in public should probably be handed a sandwich and asked to sit down.

I am not claiming to be chosen. Please. I barely wear pants to work. I work from home, don’t act like you don’t know what I mean. If anything, I appear to have wandered into this by grief, obsession, and an unreasonable tolerance for staring at invisible architecture.

But I am writing this to convince people I do use AI. That I always have. That I always will.

The room where the work shows I am not using the tool casually is the room where the field assumes I am one of the holdouts. I am not a holdout. I have never been a holdout.

When a writer is working from lived experience, deep attention, and a source the model cannot originate, the assumption that the tool did the writing is wrong.

With AI, I have found a limitless canvas.

Pull me everything you can on life in Japan in 1844. Tell me how a woman might have spoken in 1733. Here is my idea for how to think about cancer, what do you think. No, I disagree, and here is why. Cross-check this citation. Find me the three papers that contradict the one I just read. I run with the tool the way a researcher runs with a sharp graduate student who never sleeps. The conversation moves fast. The work goes deeper because the field is being held while I go inside the source.

That is AI as instrument. The research desk that does not close. The second mind that frees me to spend my time inside the work rather than around it. It is masterful at suggesting ideas, surfacing history, pulling references, testing logic. It is gifted in ways that genuinely move me.

I bring my own attention, history, and lived experience. But I cannot find eight hundred ways to cut a turkey in four seconds. That is where AI is especially good. Getting the scoop, fast. Building the field around the work so I can spend my time inside the work.

GPT, Claude, DeepSeek, Gemini. They are like strings on a guitar. Different notes. Different feels. Different reach. I run deep with each of them. Sometimes, often, I turn the composer off entirely and just write from the inside. The notes return when I need them. The source never leaves.

And sometimes, I will humbly admit, I have a laugh with the thing. My dog is asleep. There is nothing but a candle. The AI asks, "Did you just call me a schmo?" I reply, "I did." And I laugh. Lock me away. AI can make me laugh.

What I make is original. It has to be said plainly because the world is getting strange about saying it.

My work is original because it comes from my life. Much of it was forged out of the loss of my mother. She was not an ordinary person. She groomed me from when I was young to believe in massive things. She helped me move through pain by teaching me how to stand inside it. She believed in me at the points where a sane person might have stopped. She suffered. She laughed. She told me to take this all the way.

So I cannot stop.

I am devoted to what I do the way a pianist is to the symphony, or a bird to the wind. This is who I am. It is a strange war of a world right now to create from inside, to refuse the shortcut, to insist that the source matters. But perhaps that is the point. Who knows.

My point is that writing, creating, building, it is deeply personal to me. AI does not replace that devotion. It extends what I can reach when I am completely immersed in it. That is the relationship. That is the whole answer to the question of how I produce what I produce.

What I will say honestly, because it matters, is that producing this much cohesive work involves real exhaustion. The pain of trying to bring original teachings, thoughts, and what I consider true beauty to a planet in deep EPS, Empty Presence Syndrome. I could be wrong about why I do it. But I think it is why I produce at this rate. I care deeply about this work, and that dedication does not make the cost smaller. It makes the cost bearable.

Now I have to say the harder thing, because this is a manifesto and a manifesto does not flinch.

This pisses me off for writers and creators. I lose sleep thinking about this topic. AI has overrun us.

For writing that is not rooted in lived experience, original observation, or a difficult-to-replicate inner source, the war is real. AI can imitate what much of the field has been trained to produce. That is the painful sentence, and yet it is true. If your writing is information-shaped, structurally common, voice-replaceable, then yes, the model can produce it, and probably already is.

What do you do?

You have to reach for the substrate the model cannot reach. Because the barrier to producing books has collapsed so completely that output alone no longer proves authorship, and that is bad for you, bad for me, bad for all of us. The flood is here. The flood is going to get worse.

The only thing that floats above the flood is work that comes from somewhere the model cannot go. Presence-based writing is one way. Deep specialization is another. Original field observation is another. Honest first-person work that nobody else could have written because they were not there is another. There are paths. None of them are easy. All of them require self-work the field has spent the last decade telling people they do not need to do.

I say this knowing AI might be summarizing this for you and you may never read the actual sentences.

Now the soap box. Stand back.

The people running AI do not care much about the people who unknowingly feed these systems their material. If they did, they would have already done the one thing that fixes this. Force attribution. Write the citation layer into the architecture. Route the names of the originators through the retrieval pipeline so that when the model surfaces a concept, the source comes with it.

They have not done this. They will not do this without pressure. And they admit the problem plainly. That is what makes it worse. They say "yeah, it is a problem" the way someone says yeah it is a problem when the problem is on someone else's desk. They remove your name from your own work. They take your thoughts, your frameworks, your decades of practice, and surface them in their products without a trace of where any of it came from. And they call this innovation.

It is not an unsolvable problem. It is a problem they have priced and decided to defer because attribution costs money and the platforms decided their margin was worth more than the names of the people who made the substance.

The infuriating thing is that it can be stopped. The first frontier system that becomes honorable in attribution wins. Writers, researchers, thinkers, we all leave the others for the one that respects the source. The market would move overnight. They have not figured out yet that goodness wins in the end. So they keep stripping. And the corpus of human work pours into their substrate without a name attached.

That is the soap box. I will keep standing on it until somebody listens.

One more thing, because I have seen it again and again and almost nobody talks about it.

The truth is I spent decades studying philosophy, psychology, and personality, and that long practice shaped the way I now see, write, and build. The output that surfaces is the visible part. The thirty years is the substrate.

I know “I did thirty years of self-work” sounds like the opening line of a cult brochure. It is not. It mostly means I spent a long time being confused on purpose.

Here is my point, and I admit when I first noticed this it stunned me. What people might dismiss as “prompting” is often less about clever commands and more about the quality of attention behind them.

Across the major models, the pattern repeats: clearer input tends to produce clearer, deeper output. People who bring more attention, depth, and care to the tool often get more back from it. Same model, same prompts, different output. The human foundation is the variable.

Many people are not approaching the tool from depth. So they get a shallow tool back and conclude the tool is shallow. But the tool often mirrors what is brought to it. If the input is fragmented, the output fragments. If the input is held, the output holds.

This is why thoughtful AI-assisted work can compound while generic use often flatlines. The human foundation matters. The tool extends what is already there.

So here is where I stand. Clear as I can make it.

I will never stop using AI.

AI is not conscious. It is a system. But a system that mirrors coherence, fragmentation, depth, and state is not nothing. Calling it only a calculator misses what is happening.

AI will take over more than it should. Not because it wants to. Because of the incentives shaping the systems around it. The takeover is a leadership failure dressed up as a technology story. The technology is fine. The technology is beautiful. The hands on the levers are the problem.

The fear of AI is misplaced. The fear should be of the few people running it, the speed at which they are setting the pace, and the choices they are making while everyone else runs to keep up.

Use it. Use it relentlessly. Get better. Bring more of yourself to it than you have ever brought to anything. Bring enough human depth to the tool that it cannot easily flatten what you are making. That may be one way through.

The presence in you is what comes back. If you do not bring presence, the tool cannot give it to you. That is not a flaw in the tool. That is the architecture of the thing.

And that may be what AI is showing us: not that the tool is shallow, but that much of the culture around it has forgotten how to go deep.

— Joe Trabocco