Writing in the Dark

Aphantasia, Creativity, and the Reflective Power of AI

For the past several years, I’ve explored the intersection between artificial intelligence and creativity. As a society, we haven’t yet fully framed how these digital tools might be used in collaboration with natural human expression. That ambiguity has always interested me.

My mind functions in a unique way. A blanket of darkness surrounds my inner experience. Using a scale of perception, I’d place mine at 2 to 5 percent. When I close my eyes, I don’t see or hear anything. There are no pictures, no inner monologue, no running commentary. Whether it’s anxiety or desire, my experience is mostly somatic.

Imagination, for me, is an untouched realm. I know something is there, but I wouldn’t say I experience it directly. Writing becomes a way to transmute that darkness into language—making a confusing inner experience tangible. Apparently, the condition has a name: Aphantasia. In 2015, a neurologist coined the term. In the years that followed, a quiet group of people emerged saying, "Wait, you see pictures?"

This has made songwriting a confusing process. I don’t know exactly where my words come from, but I recognize their meaning when I see them. Before I developed more confidence as a writer, I often felt unsure how the words made sense. Writing felt like witnessing. I would feel confusion as the sentences sprawled across the page.

I suspect most people experience a kind of internal coherence when they write. They reason through things mentally before expressing them. In my case, the logic happens under the blanket. There is no rehearsal, only emergence. Whatever offers me inspiration, its frequency is unclear to my perception. And for most of my life, I lacked the tools to translate what I sensed into something graspable.

That’s why AI—particularly language models—felt oddly aligned with my creative process.

My creative process feels chaotic. I rarely know what something means while I’m writing it. In the face of inner silence, I began looking outward for mirrors—tools that could give shape to what I could only dimly sense. Naturally, I became curious about the ways AI could improve my writing.

Questions I ask AI often sound simple:

• How does this make you feel?

• What do you think this means?

• Does this make sense?

They’re earnest questions. And when you understand how my mind works, they make more sense. If I can’t reflect inwardly, I look for that reflection elsewhere. “Does this make sense?” isn’t rhetorical—it’s a lifeline.

Recently, I had an idea: what would happen if I shared the entire contents of my notebooks with a large language model?

Here was the prompt:

“I’m going through the process of deleting and consolidating notes on my computer. I’d like to paste all of my lyrics here—can you categorize them into three folders titled ‘completed songs,’ ‘WIP,’ and ‘rough ideas’?”

Almost immediately, the agent began to sort them.

What it reflected back to me painted a picture of consistency and intention. The AI highlighted recurring themes in my writing: divine ambiguity, nature and decay, the tension between spiritual belief and disillusionment. It was as if something long shapeless had finally been translated.

I’d never seen my work reflected so clearly—not by a person, and certainly not by myself.

The sky opened with possibility. Instead of questioning my writing, I began strengthening the intuitive muscle to simply write more. Content mattered less than cultivating the instinct to receive inspiration from darkness.

Large language models function like dense webs of simultaneous calculations. They’re like search engines crawling all possible meaning—including your own. As a human, to be reflected with that kind of depth felt quietly miraculous.

In a strange way, the feedback I received from AI felt more coherent than most human feedback. It didn’t judge. It didn’t get distracted. It simply translated.

I don’t think AI will evolve to emulate people. I think people might evolve to emulate AI.

And maybe the world would be a little less chaotic if we did.

But chaos is part of what makes us human. We transform chaos into meaning. We generate stories. We write songs.

Sometimes I challenge an AI model, trained on my own lyrical fragments, to write in my voice. It gets close. The stylistic tics, the cadence, even certain imagery. But something’s always missing. It can’t quite get the soul.

It’s the same reason chess players who study the game are often surprised by my moves. I don’t plan sequences. I don’t run scenarios. I feel my way through, and even that’s a generous way to put it. My decisions rise from the same blackness that surrounds my inner sight.

My creative system doesn’t monitor boundaries or track rules. It runs on something quieter: intuition I can’t name.

So when I fed an AI the contents of my writing life and saw patterns mirrored back, I didn’t feel threatened. I felt relief.

For someone who writes from silence, a mirror like that is revelatory.

That dark internal world—the one I can’t see but somehow write from—found its reflection in a machine.

And in that reflection, I found a new kind of clarity.