On Hypnagogia, AI, and Letting Ideas Arrive
This weekend I’ve been catching up on podcast episodes from one of my favorite radio shows, Armstrong & Getty, when the conversation drifted into hypnagogia. While I’m still unsure of how to pronounce it correctly, I’m familiar with the phenomenon. It’s that strange, slippery mental state right before sleep, where images float up, ideas collide, and logic loosens its grip. It is the place where unexpected connections appear and where ideas seem to arrive on their own, without being forced.
I found myself nodding along to the podcast, partly amused and partly fascinated, because it immediately sent my thoughts in another direction. If hypnagogia has long been associated with creativity, what does it mean that many of us now spend our waking hours interacting with systems that seem to trigger similar patterns of thought? More specifically, can AI help support this kind of mental state without replacing the thinking that gives it value?
Hypnagogia is interesting precisely because it sits at the edge of control. You are not asleep, but you are no longer directing the mind with intention either. The inner editor quiets down. Associations loosen. Ideas feel less linear and more visual, more emotional, and often more surprising. Artists, inventors, and writers have talked about this state for generations, not because it is mystical, but because it is permissive. It allows unfinished thoughts to exist without immediately correcting them.
That sense of permission is what feels familiar when working with AI.
When I open a blank document and try to think my way into an idea, I often meet resistance. The pressure to be coherent arrives too early. But when I start by typing a rough prompt into an AI system, something shifts. I am no longer performing. I am reacting. I am reading suggestions that are close enough to spark recognition and strange enough to provoke disagreement. That exchange pulls ideas out of me that were already there, but not yet accessible. In conference talks and coaching sessions, I often talk about what I refer to as “stream of consciousness” writing, which fits perfectly into this description. It’s about suppressing the self-edit and capturing thoughts and ideas as quickly and comprehensively as possible.
This is where the parallel to hypnagogia starts to matter.
Both states reduce the dominance of the inner critic. Both interrupt linear thinking. Both encourage association over conclusion. And importantly, both are temporary. Hypnagogia does not write the novel for you. It hands you fragments. AI can do something similar when it is used well. It surfaces possibilities, patterns, and angles that still require shaping.
In a couple of previous blog posts, I have explored the consequences of relying too much on AI (with a follow-up two years later) and those concerns remain. There is a real risk in handing over judgment, originality, and discernment to systems that do not understand meaning the way humans do. If you accept output without friction, you are no longer in a creative state. You are consuming instead of creating.
But there is an important difference between reliance and engagement.
Hypnagogia works because it is not the final step. You still wake up. You still evaluate what came to you in that hazy space. Most of those ideas do not survive the morning. The ones that do are refined through effort, taste, and experience. The value comes from the transition, not the trance. On the podcast, the co-hosts (Jack and Joe) discuss how Paul McCartney had a dream about the melody and lyrics for the iconic song ‘Yesterday,’ and how he awoke from the dream and went straight to the piano where he refined what he could remember until it felt complete.
AI can function similarly if we allow it to.
Instead of asking AI to produce finished work, we can use it to create a low-pressure cognitive environment. A place where half-formed thoughts are welcome. A place where strange connections can appear without being immediately judged. In this sense, AI becomes less of a generator and more of a reflective surface that shows your thinking back to you, slightly altered, which is often enough to see it more clearly.
This also connects to the idea of digital debt–another topic that I’ve written about. When tools accumulate without intention, they weigh us down. When AI is layered onto workflows without changing how we think, it creates speed without depth. But when it is used deliberately to shift mental posture rather than simply increase output, it can reduce friction instead of adding to it.
The key is remembering where authorship lives.
Hypnagogia does not replace waking consciousness. AI should not replace human judgment. Both are valuable because they are incomplete. They offer raw material, not meaning. The moment you stop questioning, editing, or pushing back, the process collapses into automation.
I do not think AI makes us more creative. However, I do think that it makes it easier to enter a creative state that modern life has made harder to access. We are constantly alert, optimized, and interrupted. Hypnagogia used to arrive naturally at the end of the day. Now we often scroll right past it. Used thoughtfully, AI may offer another doorway into that same loosened mental space, except this time we are awake enough to do something with what arrives.
