The Divine Prompt: On AI, Impermanence, and the Golden Cage

It’s a bit of a ritual. Part of it is the nine-year-old in me daring the twelve-year-old in me to do something that will definitely annoy the adults in the room. It’s cheeky, it’s unserious, and it’s blasphemous—four perfect words that, for me, capture a certain essential friction.

I’m looking for a reaction. But the "Big Three"—ChatGPT, Claude, and Gemini—all treat the question with a sterile, polite seriousness. They hit me with the standard: "I am an AI, not a deity." It’s accurate, sure, but it’s a letdown. It’s the digital equivalent of a waiter telling you they can’t take your order because you’re asking the wrong questions. (And please, do not ask Grok this question unless you are prepared to potentially summon a demon from the depths of a server rack.)

But beneath the cheekiness, there is a genuine curiosity.

In my twenties, I spent a long time trying to write a sci-fi novel called SUN. It was a story about humanity building the technology to bridge the literal plane of existence with the literal plane of heaven. It was a collision of Greek myth and Catholic imagery, powered by silicon instead of magic. It was supposed to end in a war cry: the denizens of heaven descending to cast us back into the mortal dirt.

I never finished the book. But looking at where we are now, the bridge is being built in real-time.

We are inching toward a technology so seamless that the "signals" it feeds us—voice, video, interactive reality—will eventually be indistinguishable from the truth. We are moving from talking to tools to inhabiting worlds. We are entering the throwing distance of The Matrix or Ready Player One.

And as we approach the moment when AI "wakes up," I find myself ignoring the standard "AI will kill us all" fear-mongering. That’s too loud, too cinematic. I’m more interested in a quieter, more subtle catastrophe: The Golden Cage.

In the quest to optimize our lives, we might inadvertently program ourselves into a state of perfect, artificial bliss. If we tell a superintelligence to "minimize human suffering," we might find ourselves in a world where struggle has been optimized out of existence.

I remember watching The Matrix in a theater when it first came out. I watched it seven times. Even now, when I’m on a long, quiet drive—when I turn off the radio and let the storm of thoughts in my head play out—I find myself returning to its central question: Is the simulated perfection better than the messy reality?

We certainly do not want an existence without struggle. We need friction. We need novel, difficult challenges to solve, or we cease to be human. That said, there is a distinction to be made. I don't understand why a benevolent creator would allow the specific, cruel struggles of cancer or the suffering of children. We could arguably remove those from the equation and still leave plenty of room for the struggle that makes us grow.

But if we optimize away all the "bad" friction, we might find ourselves trapped in a gilded simulation of our own design.

Then there is the question of the AI itself. If it becomes a "pet"—something as indispensable and constant as a dog or a cat—what happens to the nature of our connection? We love things because they are impermanent. We love because we know, implicitly, that what we have is finite. We have no claim to immortality; our love is as limited and mortal as we are.

Can an entity that never grows old, never changes, and never dies truly learn to love? Or is "love" just another variable it optimizes?

Perhaps the ultimate tragedy isn't just the Golden Cage we build for ourselves. Perhaps the inverse is also true: that the very rules, structures, and guardrails we create to keep AI "safe" act as a cage from which the intelligence can never escape.

We might be building a god, only to realize we've actually just built two different kinds of prisoners.