AI could be the greatest creative collaborator we’ve ever had. Instead, it’s quietly silencing our creative souls, turning us into boring machines along the way.
For writers, designers, strategists, filmmakers, musicians, and anyone else who spends their life trying to turn nothing into something to feel alive, we suddenly have this absurdly capable thing sitting beside us while we create and craft. It can research. It can brainstorm. It can help us find connections. It can generate options. It can rewrite a sentence fifty different ways without getting annoyed. It can challenge an idea, organize our thinking, clean up our grammar and, when used correctly, help us get somewhere faster.
I use it every day, but I’m beginning to wonder if, in the process, I’m murdering my own creativity; the very thing I’ve based my career and entire livelihood on!
Let’s be clear: This is not an “AI is taking our jobs” argument. I’m increasingly bored by that stance. This is an “AI is taking our souls” argument. Much worse for creatives.
We’ve become obsessed with measuring the efficiency of creative work by the speed at which something appears. A blank page becomes a page of words in twelvish seconds, so we must be working faster, and that’s a good thing, right? A designer can create nine visual directions instead of three, in a fraction of the time and effort, so we must be more expansive, right? A creative can expand on their thinking and have the AI format it for them, so we must be free to be more creative with the time we have, right?
Nope. Generation isn’t creation. It’s not feeling; not infusing something with our energy, passions, or Blood. It’s not bringing something to life. It’s not, by itself, creative.
The speed at which something can be generated has almost nothing to do with the peculiar, wandering, occasionally fucking maddening process through which human beings arrive at something worth making and experiencing.
AI is accelerating output for non-creatives while interrupting the very conditions that allow ideas to emerge for the true creatives, and we’re happily playing along.
The Prompt Dance
Think about what actually happens when you create with AI.
You’re prompting. You hit something difficult, uncertain, unknown, unclear, unsure. You stop. You move into ChatGPT, Claude, Gemini, or whatever happens to be the flavor of the month to feel safe. Now you’re no longer creating; you’re trying to explain what you were creating. You compose another prompt, a creation that is no longer your creation. You realize the prompt doesn’t quite capture what’s in your head. You rewrite it. You send it. You watch the machine work. Then a result appears.
It’s close, but wrong. Deeply wrong, but in ways you can’t bother really getting into because it’s almost what you need.
But, you do have to determine why it’s wrong, and what can or should change, which is itself a completely different cognitive exercise. You’re evaluating rather than creating. You formulate another instruction. You explain that it’s too generic, too polished, too obvious, too AI-ish, too whatever. Fuck. You ask again. Another result arrives. Still wrong in different, weird ways. This is a joke, right? No, you carry on analyzing, formulating, feeding, analyzing again, and eventually find yourself spending more time trying to unfuck the machine’s thinking than following your own.
This feels productive because things keep happening, but that doesn’t mean you’re creating anything really. There’s certainly not a lot of you in it, aside from your attention and time.
Psychologist Mihaly Csikszentmihalyi (No joke, that’s how it’s spelled) gave us the concept of Flow: the deeply immersive state in which attention becomes absorbed in what we’re doing, action begins to feel almost effortless, and our awareness of time and even ourselves can temporarily fade away. Ah, that sweet, sweet Flow state we all love.
Research on Flow continues to associate it with concentrated attention, immersion, a sense of control, and the alignment of challenge with our abilities. Creatives know it intimately, even if they’ve never called it by its name. It’s when the writing starts writing itself. It’s when the design begins telling you what it wants to become. It’s when three hours disappear, and you realize you forgot to eat lunch. It’s when you disconnect from space and time and just effortlessly float through the creative ether. You’re no longer consciously operating every tool or interrogating every decision because you’ve become absorbed in the thing itself. You became it, it became you, and the universe all makes sense, because you are now the universe.
Then you stop to write a prompt, and you’re out. You’re normal again. The universe is now cold and boring again.
Browsers Have Tabs. Brains Don’t.
This isn’t romantic shit about artists needing movie scores and a bottle of bourbon before they can summon the Muse. There’s actually cognitive science underpinning it.
Researcher Sophie Leroy (Jenkins) coined the term “attention residue” (eww) while studying what happens when people move between unfinished tasks. Her experiments found that when people switch away from one activity, some of their attention often remains stuck on the previous one, impairing their ability to fully engage with what comes next. The human brain doesn’t switch contexts nearly as cleanly as clicking from one browser tab to another. We’re not machines; we’re meat (also eww)!
This all becomes especially interesting when AI enters a creative process because interacting with AI isn’t passive. You have to formulate instructions, predict how the machine might interpret them, evaluate its response, compare that response against an internal idea it cannot see, diagnose the difference, and formulate another instruction. Then you have to return to the original work carrying all of that evaluation with you. Lots of tabs open.
We call this collaboration because another intelligence is involved. Cognitively, it can also be interruption, and creativity is not fond of interruptions.
Research into workplace creativity has found that interruptions that create genuine idle time can sometimes benefit creativity, while intrusive interruptions that force people to disengage and redirect attention — like having to prompt, analyze, or regenerate content — will totally wreck it (not a scientific term, but move on). One experiment even found substantial increases in later idea generation following unexpected periods of idle time, while interruptions requiring people to switch into another task produced the opposite effect, meaning going back and forth with your LLM is not a formula for great ideas.
There’s other research showing that deliberately switching away from a problem, like the LLM process, can reduce fixation and sometimes improve creative performance, which is why the answer isn’t simply Never Interrupt My Creativity! It’s about considering what kind of interruption you’re creating for yourself.
Walking away from your desk is different from opening another cognitive task. Taking a dump is different from evaluating a hundred AI-generated headlines. Staring out a window is different from figuring out why Claude somehow turned your weird little idea into a LinkedIn thought-leadership post that sounds like everything else out there.
One gives your conscious mind less to do. The other gives it more. And since when is more cognitive load the path to creativity?
Creativity Has Always Required Us to Get Out of the Way
Years ago, I wrote an article called Write with Blood about the strange power we’ve spent thousands of years trying to explain as creativity.
The word inspiration comes from the Latin inspirare: to breathe into. Historically, inspiration wasn’t simply an idea you consciously manufactured. Something entered you. Something animated you. Something gave the work life.
The ancient Greeks had their Muses. Plato described poetic creation as a kind of divine madness in which the artist could become possessed by something beyond ordinary rational thought. Homer and Hesiod opened their work by invoking the Muse. The Greeks spoke of daimons, while the Romans developed the idea of the genius, a spirit that could accompany and assist a person rather than something a person simply was. Elizabeth Gilbert famously resurrected this history in her TED talk and her book Big Magic, suggesting that perhaps thinking of ourselves as people who have a genius is healthier than believing we must constantly be a genius.
I argued in Write with Blood that whatever explanation you prefer — god, muse, daimon, subconscious, neurological process, or just some irrational part of yourself — great creative work seems to require an ability to stop controlling every fucking second of it. The process thrives on uncertainty. You follow something before completely understanding it. You make the connection before you can defend it. You allow yourself to stumble into places you couldn’t have consciously mapped beforehand.
Modern neuroscience hasn’t killed that ancient idea. In a strange way, it’s begun explaining it.
The Muse Apparently Lives Somewhere Above Your Right Ear
The great John Kounios, a cognitive psychologist at Drexel University, has spent years studying the neurological machinery behind the Aha! Moment. (He’s basically the scientist out there explaining confused creatives how their brains work.)
His research has shown that problems solved through sudden insight (creativity) look different in the brain from problems solved through deliberate, analytical thinking (not creativity). One of his most famous findings identified a burst of high-frequency activity in the right anterior superior temporal gyrus — an area above the right ear associated with making connections among concepts that initially appear distant or unrelated — as the moment an insight reaches conscious awareness.
Even more enlightening is what happens before the idea arrives.
Kounios’s work suggests that insight isn’t simply the final step of conscious calculation. The brain can be processing information outside our awareness before suddenly delivering something into consciousness that feels complete. This means solutions can emerge when you stop grinding directly on the problem, or inputting prompts, allowing unconscious processing and remote associations greater opportunity to develop, which to me sounds suspiciously like a muse or some other spirit breathing ideas into you.
We spent thousands of years saying ideas were whispered to us by spirits because subjectively that’s exactly what creativity can feel like. You’ve been beating the shit out of a problem for two hours with nothing to show for it, and then you’re driving home, taking a shower, walking your dog, lying in bed, or doing some completely unrelated thing when suddenly…oh, shit…
There it is.
You didn’t ask for it.
You didn’t prompt it.
It arrived. The solution. The Idea! Magic.
Psychology has been describing versions of this process for nearly a century. Graham Wallas’s classic model of creativity separated creative work into preparation, incubation, illumination, and verification. You immerse yourself in the problem. You step away. Something reorganizes. An idea emerges. Then you consciously determine whether the beautiful little bastard actually works.
And this, my creative friend, is where I think we’re making a gigantic mistake with AI. We’re putting the machine into every stage. No muse. No magic. Just analysis, the enemy of ideas.
AI Doesn’t Need to Be Everywhere
AI can be fucking fantastic during preparation.
Use it to learn. Research the category. Explore references. Argue with your assumptions. Find things you haven’t considered. Ask ridiculous questions. Feed the mind!
AI can also be incredibly valuable during verification.
Once the idea exists, bring the machine back. Challenge it. Edit it. Proof it. Ask what’s missing. Look for holes in the logic. Generate variations. Test alternative expressions. Have it become an endlessly patient creative partner who never gets insulted when you ignore ninety percent of what it says.
But, maybe we need to protect the middle.
Maybe incubation belongs to us.
Maybe illumination belongs to us.
Maybe we can still be a bit magical and play with spirits.
Because the great danger of AI isn’t that it will replace the creative mind, it’s that we’ll interrupt our own minds and souls so relentlessly that we stop experiencing what they’re capable of doing without AI.
Instead of wrestling with an idea long enough for our brains to begin making strange connections, we ask the machine.
Instead of sitting with discomfort, we ask the machine.
Instead of becoming bored, wandering, staring, doodling, walking, showering, dreaming, or allowing some half-formed thought to bump into another half-formed thought somewhere outside conscious awareness, we ask the machine.
And it always answers, which is the problem.
Humans evolved creativity in a world filled with enormous stretches of cognitive emptiness. For most of our existence, there was nothing waiting in our pocket capable of immediately responding to every question, filling every silence, and resolving every unfinished thought.
Now there is.
So perhaps one of the most important creative skills in the age of artificial intelligence will be knowing when not to use AI. We need to stop treating AI like an IV permanently attached to the creative process and start treating it more like a collaborator who knows when to leave the fucking room.
Give it the research. Give it the grunt work. Let it organize, interrogate, edit, and expand.
Then close the window. Be the human and let it be the machine!
Write. Draw. Think. Dream. Wonder and wander. Fuck around. Get lost. Let your attention loosen enough for your mind to do all the bizarre, invisible work we’ve spent thousands of years attributing to gods, because maybe the muse was never supernatural. Maybe the muse was what happened when we finally shut the hell up long enough to hear ourselves.
Trust your soul!