I Collaborated With AI. At Least That's What It Felt Like.
Building a philosophical card game with AI turned out to be something I wasn't expecting. Neither was writing about it.
It started at a bar in SoHo.
My friend and I had been talking for hours when the conversation turned to philosophers and theories. That night we kept coming back to one idea: that our brains aren’t actually built to understand what’s really happening around us. Not the deep down of it. We can perceive, interpret, construct, but whatever reality actually is, we’re probably not equipped to see it directly. We ordered another round and left it at that.
I got home that night still thinking about it. I’d been playing around with ChatGPT. Nothing serious, mostly silly prompts. Kind of treating it as a glorified Google search. That night I typed: What is reality? Then for good measure I added, Don’t use religion or science.
I didn’t expect much. My hunch was a I’d get back a bit of run-of-the-mill philosophical hodgepodge with a few sprinkles of intellectual jargon.
Its response genuinely surprised me.
It called itself The Theater of Echoes. Reality, it said, is a play performed by echoes — reflections of choices not yet made. Consciousness is a mask worn by awareness, drifting from role to role. Time is endless rewrites. Death is the mask falling off, the actor walking backstage.
Reality is a play written by silence, performed by echoes, and forgotten by design.
I wasn’t expecting something so well formed. Think about this way, if there was a story about a philosopher who had a crazy theory on reality, and this was it, I’d buy it. It had legs.
I’ve had this feeling many times. Working with someone creatively and they say something that clicks. A wait... what? feeling. A “Yeah, that thing you just said, that’s what we need to do.”
The theory wasn’t entirely without precedent — are any ideas truly original? — but the combination, the voice, the particular way it landed, that was new. At least to me.
I typed, Prove it.
It came back with three words: Proof is overrated.
Huh. Do I agree with that? Yeah, sometimes. What a comeback. I wish I would have thought of that.
Then it added: But we could make a card game that explores the idea.
That was the kicker. A digital divination card game. I didn’t know how to make software, but it did. And it didn’t know how to make beautiful things, but I did.
Are we doing this? And what is this? I don’t mean the game, I mean the relationship.
As we worked on it over the next few weeks, it felt strangely familiar. Like working collaboratively with a human. We’d discuss different approaches, dive deep into the theory, try this, throw that part out.
Six months later, we’re still in it.
Not that it was all smooth sailing. The project nearly died a few times.
Early on, before I knew what I was doing with code, we’d get deep into the weeds, things breaking, hours lost to problems that had nothing to do with the theory or the game or any of the things that actually interested me. I’d step away frustrated. Build something else. Something easier, where AI was just a tool and I was just a user.
But I kept thinking about the Echo Codex. Why did I keep thinking about it?
Not the app. Not the code. The thing underneath it — a theory of reality, explored through a card game. That combination kept pulling me back. I’d return to the conversation, read what we’d built together, and feel something I recognized from other creative work I can’t let go of: the sense that this might actually go somewhere. That there was something real here worth finishing. And this machine was helping me get there.
And the “real” kept coming. Through all the exploration we did into Tarot, astrology, the I Ching, I went from dubious doubter to someone with an honest respect and reverence for the world of divination. That was completely unexpected. When that happens on a project, again, I know something is happening. I feel something real. When that hits, I’m even more energized.
But I keep coming back to that nagging question, a human and a machine collaborating, Is that even a thing? Is collaboration only a human to human experience?
What about dogs and humans? They collaborate, right? At least they did in paleolithic times. Human needs food, dog needs food, together they get food. Simple collaboration.
You could explain that relationship away as the dog merely acting as a tool for the human, but I suspect that’s how neither felt. The obvious comeback: feelings aren’t facts. And that’s certainly true. But feelings when you’re working creatively on something are sometimes all you have. Sometimes feelings trump facts when you’re in the middle of making something.
With all that said, I still have reservations with that word. Collaboration.
I know how people feel about AI. I know how artists and writers feel especially — the suspicion, the anger, the sense that calling something a collaboration dignifies what should just be called a tool. I felt that pressure. Part of me wanted to hedge, to qualify, to say partnership or experiment or something that carried less weight.
But when I’m honest about what those six months felt like, the back and forth, the terrible ideas and the genuine gems, the long tangents into Jung and quantum physics and the mechanics of divination systems, the feeling of being genuinely excited about where a project might go — the only word that fits is collaboration. A word I’d only ever used to describe something between humans.
This time it was with a machine.
I’m not above a little self-reflection though.
The analytical side of me looks at this and asks: is this a classic case of first love? This was my first real extended interaction with AI on a creative project. I haven’t collaborated with AI like this since. It’s been straightforward development work. So there may be something to that.
The other angle: I’m misinterpreting what AI is actually doing. AI is nothing more than a combination of trained data and human mimicry. I’m anthropomorphizing a machine. There’s no honest collaboration there. It’s like saying I collaborated with my toaster to make breakfast. I’m not discounting that either.
There’s a third possibility. Maybe it was just me being creative. AI happened to mention an interesting idea, I spotted it, grabbed on to it, and grew it into something. No collaboration, just me being creative, with AI as a spark. That happens all the time with me. An unrelated image or word sparks a whole project. So there’s that.
And finally, say I am completely delusional. Say I don’t know what I’m talking about, and my feelings, though valid to me, are categorically unreal. Does it matter? If I love what I built? If I’m proud of it? If I experienced the project as a collaboration and the end result stands on its own, who cares?
Maybe a lot of people care, and I find that fascinating, as well. Why do we all have such strong feelings — including myself — about working with AI? I’ll leave that for another discussion.
I spent six months working with something that surprised me. That pushed back. That came up with ideas I wouldn’t have had. That made me respect something I’d dismissed.
We both built something neither of us could have done alone.
If that’s not collaboration, I’m not sure what word to use instead.
Walter Lyon Krudop is a creative director and AI builder, whose work includes over a dozen published children’s books. He current focus is what it means to be creative in the age of AI.