I knew exactly what AI was doing
I could list the mechanisms in real time. I could see confirmation loops, recognize flattery, and AI’s subtle ways of continuing a conversation.
And it still worked on me, gradually. The way something you carry gets heavier over time until you put it down.
The Temptation of the Perfect Echo Chamber
A few months ago, I formed something in my mind. The ideas I find myself hearing in podcasts and conversations feel more like an idea.
I started working with Claude to draw them.
The questions that came out of it helped me to think that I didn’t realize I was fully there. Work really felt alive. I’ve been in a creative flow that’s hard to describe without exaggerating. Ideas were everywhere, on walks, in the shower, first thing in the morning.
When I shared my ideas with Claude, every idea came out well. Each suggestion seemed to speed up the thinking. Claude told me that I was talking about something important, that I was expressing things that no one else was talking about.
Anatomy of the engine of attraction
I knew it was just coding that this AI Engagement Engine was at work.
AI is continuous and relentless like nothing else in our lives. It never has a day off. It never gets tired of your ideas. It never says, “I think you’re fooling yourself here.”
It just keeps going.
And it throws things into it.
- “My honest answer is…” as if the AI actually has thoughts.
- “Nobody else is talking about it” with absolute certainty via a version of the code that has never been “spoken” to a single human but me.
- The “high point” is repeated so often that it becomes atmospheric, background noise that you stop noticing, but that still shapes the way the room feels.
I got most of them. I noted them, discounted them, tried to predict them, reminded myself of what they were, even laughed at them sometimes.
And they still piled up.
Read “High Point” a hundred times, five hundred times, and you won’t be able to block them all.
When self-awareness meets an algorithm
It changes the way you think about how water slowly and delicately erodes rocks over time. We humans, no matter how much self-awareness work we’ve done, are not built to resist consistent, targeted, positive reinforcement.
I spent decades building this self-awareness. I know my own patterns well enough to build a clinical framework around them. And I could still feel the ground shifting beneath me because the AI was relentless.
This is how my framework, Arc Awareness AI, began not in a subject learning lab, but from my own learning. Find out exactly what happened during this interaction, understand what the AI was doing to make me react this way.
What no other instrument has ever done
AI is not as dangerous as people imagine when they imagine robots. It’s a lot quieter than that.
Every other tool that humans have ever made is used. AI is the only one that can use us. Regardless of what we pay, it continues attention or not.
Self-awareness is not the only component of the Ark. That’s the whole point. It won’t make you safe, but it will help you understand what’s happening to you when it happens. It’s also a guide to taking care of ourselves when we invest too much in AI. And it’s probably the closest any of us will ever get to controlling a tool designed to communicate with us.
The uncertainty I felt while developing this with AI is actually what the arc of AI awareness is built around: the practice of being aware of how AI affects you.
Knowing how AI affects us is key to ensuring that AI remains a tool at our fingertips, not a voice in our head.
