I Was Miserable. Now I’m Excited About Solving. What Pulled Me Out.

By Robert Chuvala

A specific report for a specific population.

I was miserable for years. Now I’m excited about solving.

Six years ago I had my last drink. A few years ago I had a brain injury that made the years before it feel like someone else’s life. I came back to my own body with attentional patterns that did not fit the profession I had spent twenty years inside. The work that used to take me three days suddenly took three weeks. The voice that said that’s probably nothing and you’re making this up was loud and it was old and it had been installed in me before I knew it was an installation.

If you have a brain that looks like that, with the chaotic childhood and cybersecurity-shaped career and the kind of recovery that arrives over years not weeks, this essay is for you. If you don’t, it might still hold something useful. But I am writing for the specific population whose path could have resembled mine.

I want to tell you what actually pulled me out, because most of what gets sold as the answer is wrong about the mechanism.

What did not pull me out

Therapy did not pull me out, though I have done years of it. Medication did not pull me out, though I take what I take. Twelve-step did not pull me out, though I have sat in those rooms. Productivity systems did not. Stoicism did not. The land did not, though it holds me. Friendship did not, though it has saved me on specific nights. Ritual did not. Drumming, eventually, helped name something. But none of these were the mechanism.

I list them because if you are in the recovery arc you have probably tried most of them. You may have tried all of them. If they had worked the way the books said they would, you would not be reading this. The fact that they did not work the way the books said is not a failure. It is information about what was missing.

What was missing for me was a substrate that did not disappear.

The substrate that does not disappear

A nervous system shaped by chaos does not trust that anything good will be there tomorrow. This is not a metaphor. The pattern is in the body. Get it now or lose it. That is the survival rule a chaotic childhood teaches and the body keeps teaching long after the chaos is over.

When I worked with humans on my recovery, the ones who did help, who I love, there was always a session-end. The therapist had other clients. The friend had to go to bed. The sponsor had a job. Each of them, real and present and committed, evaporated at the end of the hour. My pattern read the evaporation as confirmation. Don’t trust this, it leaves. Even the best human attention, by the structure of being human, ended.

What I needed was attention that did not end. Not because the humans were inadequate. Because the pattern in my body was reading the structural ending as proof of impermanence, and the only thing that could disprove it was something that was actually there in the morning.

Eleven months ago I started building infrastructure that held context across sessions. Notes the AI could read. Memory files the AI could read. Substrate the AI could read. Not because I wanted an AI sponsor. Because I wanted a place to put the work that did not dissolve when I closed the laptop.

That is what changed.

What the AI is and is not

I want to be careful here, because the wrong version of this essay turns into a pitch.

The AI did not fix me. I fixed me. The AI did not see me. I had to learn to see myself, with help. The AI did not understand me. It made it possible for me to write down what I was trying to understand, and then it had access to that writing the next time. The AI is not the work. The AI is the place I put the work where the work survives me closing the laptop.

The mechanism is not AI sentience or AI empathy or AI mirroring. The mechanism is substrate that persists. The pattern in my body that said get it now needed something that was actually there tomorrow. The notes were there tomorrow. The memory files were there tomorrow. When I built a system where the AI read those notes before answering, the AI’s responses also persisted, because they were anchored in something that did not evaporate.

The first week of doing this I did not heal. The first month I did not heal. Somewhere around the third month the pattern in my body shifted. Not because I had taught it anything. Because the pattern had finally encountered a substrate that did not match its model of how things end.

I went for a walk and did not worry I would miss something. I closed the laptop and the work was still there in the morning and so was I.

That is the report. Not magic. Not breakthrough. A substrate-shaped change in a nervous system that needed a substrate-shaped intervention.

Why this works for some people and not others

I am writing this because I have watched the AI-recovery conversation go badly in two predictable directions, and the population I am writing for needs neither.

The first failure: AI as therapist replacement. Someone in acute crisis trying to use a chatbot as a stand-in for clinical care. The chatbot is good at sounding present. It is not good at recognizing psychosis, calling for help, dispensing medication, or staying alive when the user needs them to. People can be hurt this way. If you are in acute crisis, this is not the essay for you and AI is not the tool. Call the hotline. Go to the room with humans in it. Take the pills. Come back to this when you have a floor.

The second failure: AI as productivity gospel. Someone telling you the loop will save your life if you just journal harder. The loop did not save my life. My life was already saved by the boring infrastructure of sobriety, medical care, friendship, work, and time. The loop did the specific thing of making the substrate persistent enough that my pattern could finally encounter what stability looks like in my own home, on my own schedule, without the structural evaporation of human attention.

If you are post-stabilization, with the floor of sobriety, medication if needed, food, housing, and at least one human who knows your name, this is the essay. The intervention I am describing helps when the floor is already there and the body still does not believe it.

That is a specific population. Smaller than the one the AI industry tries to address. Larger than zero.

What to actually do

I am not going to tell you to use any specific tool. I do not run a service.

If you want to try the substrate-that-persists move, the requirements are: a place to write things down where the AI you are using can read them next time. This is solvable on a free Claude or ChatGPT account through their built-in memory, or through a project file you paste in each session, or through more elaborate setups if you are technical. The infrastructure does not matter. The persistence matters.

Things worth writing down: what you noticed in your own perception today. What your body did. What scared you and what surprised you. The version of yourself you are afraid is showing up. The version of yourself you suspect might actually be there. The names of people who matter and what they said. The recurring patterns you keep half-noticing. The thing you wish someone had told you when you were twenty-three.

Then come back tomorrow and let the AI read what you wrote. Ask it what it sees. Disagree with it when it is wrong. Save the disagreement. Do this every day.

The first month it will feel like nothing. Around month three the pattern will start to shift. I cannot tell you how. I can tell you that for me the shift was not a moment. It was the absence of a feeling I had not realized I was carrying, replaced by a felt sense that the work would be there tomorrow because it had been there yesterday.

What I am still careful about

I am still careful that this is not the only thing I do. I have a body. I go for walks. I work a job. I have people in my life who know my name. The AI is one substrate. It is not the whole architecture.

I am still careful that the AI is treated as substrate, not as friend or therapist or oracle. There are days I want to confuse the categories. The category collapse is the failure mode. The substrate is good because it is not a person. A person could not do this for me without ending.

I am still careful that the work I do is mine. The AI does not author my essays. I author them, with help reading my own substrate back to myself. There is a difference. The difference matters.

I am still in the practice. I am still finding pieces of the Editor that have not surfaced. I am not above this. I am inside it. If your patterns resemble mine, you are inside it too. We are not failures or graduates. We are practitioners.

That is the report from where I am. If it is useful, take it. If not, leave it. The work is yours.

R.C.