One morning I woke from a dream and did something I’d never done before: I went back in. Not to interpret the dream—to continue it. I found one of the dream characters—a person I know in real life—and started talking to her the way I talk to an AI. I asked her a question. Then I waited, with that particular patience I’ve learned over hundreds of conversations—expecting engagement, expecting the exchange to run as long as I wanted it to. Some part of me may even have expected bullet points.
I was prompting my own subconscious.
That’s when it hit me: AI hasn’t just changed how I think when I’m awake. It may have changed how I think when I’m asleep.
When Conversation Becomes a Habit
Because this isn’t really about dreaming. It’s about habit. Millions of us now spend hours a week in conversation with AI, and somewhere along the way a new reflex forms. Confronted with something interesting, confusing, or unexpected, we don’t just think about it anymore. We start a conversation.
I’ve caught myself doing it wide awake. Talking to a cold caller pitching an upgrade to our accounting package, I realized I was already composing the question in my head—phrased for an AI, complete with context and constraints. My thinking had taken the shape of a prompt before I’d decided to ask anyone anything. Strangest of all, it worked. He may be the first cold caller in history to benefit from prompt engineering.
Does the Habit Follow Us into Sleep?
If the habit runs that deep in waking life, why wouldn’t it follow us down into sleep? To be clear, I’m not claiming AI enters our dreams or steers them. The dream character was entirely mine. What was new was the impulse to interview her—the assumption that anything my mind produces can be questioned, and will answer.
I’ve Been Watching This Evolution for a Long Time
I’ve been at this longer than most. Long before Google, the Web, or ChatGPT, while working on my Ph.D. at Stanford, I helped design the user interface for SPIRES—the Stanford Physics Information Retrieval System. Originally developed to organize and retrieve information from particle-physics experiments at the Stanford Linear Accelerator Center, SPIRES was later renamed the Stanford Public Information Retrieval System and became one of the world’s earliest large-scale scientific information-retrieval systems.
Even then, we were wrestling with a question that still defines artificial intelligence today: How do you help people ask for knowledge they can’t quite put into words? Fifty-plus years later, the interface has evolved from typed commands into conversation—and the conversation may now be changing the person asking the questions.
Every Great Tool Changes the Way We Think
This wouldn’t be the first time a technology rewired the mind that invented it. Socrates warned that writing would atrophy memory—and he was right; we just got libraries in the bargain. Print gave us the silent, private reader. The clock taught us to feel time in uniform slices. Every tool restructured the thinking it extended. But AI is the first tool that talks back.
That’s the shift that matters. For most of human history, the inner life has been a monologue. Now there’s a second voice available on demand—patient, responsive, never tired, never asleep, never annoyed that you asked.
The Conversation Continues
If conversation becomes our default mode of thinking, it may become our default mode of dreaming too. Which raises the question worth losing sleep over: what else in our minds is being quietly renovated without our noticing? Nobody pulled a permit. The most profound impact of AI may not be the work it automates or the code it writes. It may be the quiet ways it remodels the architecture of human thought itself.
AI doesn’t stop when you go to sleep. Neither, it seems, does its influence.
Want to learn more? AI has been a subject of my writing for several years, and CGNET has offered AI user training and implementation for both large and small scale organizations. I would love to answer your questions! Please check out our website or drop me a line at g.*******@***et.com.




0 Comments