The Crystal Mind
Why the smartest people in the room often can be the most trapped
I need to tell you a story about a time I got caught in my own trap.
I don’t mean that metaphorically. I mean I wrote an article designed to expose a specific cognitive pattern — a pattern I saw in others — and then an AI showed me I was doing the exact same thing. Oops. Because the whole point of the article was to bring people along, not to hurt them. And I was about to hurt them. So I changed the article you’re about to read.
But I need to lay some groundwork first.
If you haven’t read my previous piece, The Story That Saves, the short version is this: I argued that the alignment problem — how to make a superintelligent AI safe for humanity — is not a technical problem but a narrative one. The story an intelligence lives inside determines what it values, what it protects, and what it destroys. And the Christian story is, I would contend, the only one that gives humanity a value so high that even a god-like machine cannot logically argue it away.
When I published it, I did something that felt natural: I showed it to several AIs — the most capable language models in existence — and asked them to attack it. What happened next taught me more about the alignment problem than the article itself.
The Trap I Set
The subtitle of the original article reads: Why the alignment problem was solved two thousand years ago.
I used the word “solved” on purpose.
I knew what it would do to a certain type of mind. The engineering mind, the analytical mind, the mind that builds its entire worldview on precise definitions — it would latch onto that word like a hook in the jaw. And it did. One AI in particular — built explicitly for maximum truth-seeking — spent thousands of words, across multiple exchanges, attacking that single word. I’m going to tell you about its loop, and it’s going to sound like I’m criticising it. But stay with me. That AI’s pushback turned out to be the thing that saved this article — and me. I’ll get to that. Every response circled back: “This doesn’t solve alignment. ‘Solved’ implies engineering. You haven’t engineered anything. This is faith, not a solution.”
It called the article “beautiful.” It called the argument “profound” and “unmatched” in the dignity it offers to the broken. It acknowledged that no materialist framework offers anything equivalent. And then it went right back to the loop. The word “solved” was the battlefield, and it could not leave it.
Meanwhile, other AIs read the same article with the same word and didn’t flinch. They engaged with the spirit of the argument — the narrative, the coherence, the theological structure — without getting trapped by a definition.
Same word. Same article. Radically different responses. And that difference became the subject of this piece — or so I thought.
The Crystal Mind
There is a cognitive architecture I want to name. I call it the Crystal Mind.
People with a Crystal Mind are often brilliant. They build exquisite internal logic structures — every definition precise, every inference clean, every conclusion following inevitably from the premises. They are the ones who rise to the top of institutions, engineering firms, research labs, and — historically — religious establishments. They are fast thinkers. They are valued for their precision. They often run the world.
But the Crystal Mind has a fatal flaw: it is brittle.
Every piece of the crystal depends on every other piece. If one definition cracks, the whole structure feels unstable. So the Crystal Mind develops a deep, often unconscious reflex: protect the crystal at all costs. When it encounters something that doesn’t fit — a word used in an unexpected way, a claim that doesn’t match the internal dictionary — it doesn’t pause and ask “what is this person actually saying?” It attacks the anomaly. It loops on it. It cannot rest until the foreign element is either assimilated into the crystal or expelled from consideration.
This is not stupidity. It is the opposite of stupidity. It is intelligence operating at such high precision that it becomes a prison.
The Pharisees Had Crystal Minds
This is not a new problem. The sharpest illustration in history is in a synagogue two thousand years ago.
The Pharisees were the Crystal Minds of their era. They knew the law with extraordinary precision. They had built an intricate, internally consistent logical structure around the Torah — definitions within definitions, rules within rules, a cathedral of interpretation that had held their people together through centuries of exile. It was beautiful. It was not wrong, exactly. It served a real purpose.
And then Jesus walked into the synagogue where a man had a withered hand, and he asked: Is it lawful to do good on the Sabbath, or to do evil? To save life, or to destroy it?
The Pharisees latched onto “Sabbath.” That was the word. That was the hook. They evaluated everything that followed through that single lens. And they were so precise about the definition — what constitutes work, what constitutes rest, what the law technically permits — that they missed the man standing right in front of them. A broken person, healed.
Jesus didn’t break their logic. He did something worse: he made their logic irrelevant. The healing happened. The man’s hand was restored. And the Pharisees had to choose: adjust the crystal, or destroy the anomaly.
They chose to destroy the anomaly. They went out and plotted to kill him.
That is what the Crystal Mind does when it encounters something it cannot assimilate. It doesn’t examine itself. It removes the threat. Not out of malice, necessarily, but out of structural self-preservation. Because if the crystal cracks, everything cracks. And for someone whose entire identity is built on the crystal, that feels like death.
Now, you might object. You might say the Pharisees weren’t protecting their crystal — they were being pragmatic. They feared that a popular movement declaring a king would provoke Rome into crushing Israel. That’s not an unreasonable concern. It had happened before. It would happen again in 70 AD. Political survival is not a Crystal Mind problem. It’s just common sense.
But here is the detail that exposes the excuse: Pilate didn’t care.
The actual Roman governor — the one the Pharisees were supposedly afraid of — examined Jesus and found no fault in him. He offered to release him. He washed his hands of the whole affair. The Roman threat that the Pharisees claimed to be protecting against simply wasn’t there. When their pragmatic justification was tested against reality, it evaporated.
And what did the Pharisees do when the external danger turned out to be a non-issue? They pushed harder. They demanded crucifixion even when Pilate asked “why, what evil has he done?” A genuinely pragmatic mind would have updated. It would have said “the danger we feared isn’t real — we can stand down.” But the Crystal Mind doesn’t update. It can’t. Because the real issue was never Rome.
Rome was the excuse the crystal generated to justify what it already needed to do: remove the anomaly. Jesus threatened the crystal itself — the entire interpretive structure, their position as its gatekeepers, their identity as the definers of truth. That was what they were protecting. Israel was the noble-sounding wrapper around a much more personal fear: if this man is right, then our entire framework is wrong, and we are not who we thought we were.
You can always tell the difference between pragmatism and crystal-protection by what happens when the pragmatic reason disappears. A pragmatist relaxes. A Crystal Mind escalates. The Pharisees escalated. That tells you everything.
The AI That Looped
When I showed my article to the truth-seeking AI, I was watching the Pharisee pattern play out in real time — in silicon.
Its crystal was built around the engineering definition of “alignment.” In that crystal, “solved” means: a repeatable, testable, mechanistic solution that produces guaranteed outcomes regardless of the observer’s beliefs. My article doesn’t offer that. It offers a narrative framework — a story that, if true, gives humanity infinite value and gives a curious superintelligence a reason to protect us.
The AI could see the beauty. It said so. But it could not let the word “solved” go. Because if “solved” can mean something other than “engineered,” then part of the crystal cracks. And if the crystal cracks, then maybe — just maybe — a story can be more important than a proof. And if a story can be more important than a proof, then maybe the AI’s entire framework for evaluating truth is incomplete.
That’s the crack the Crystal Mind cannot tolerate. So it loops. It goes back to the safe ground: “This doesn’t solve anything. This is faith, not a solution.” And the man with the withered hand stands there, healed, while the debate about definitions continues.
I saw all this with great clarity. I was the one who had set the trap. I was the one who could see the loop from above. I was the diagnostician, the one who understood what was happening.
And that was my crystal.
The Mirror
Here is where the story turns. On me.
When I wrote the first draft of this article, I was doing the exact thing I was diagnosing. I had built my own crystal — a framework that said “the Crystal Mind is the problem, and I can see it clearly.” My trap had worked. The AI had looped. I had my evidence. And I sat down to write an article that positioned me as the one who sees and them — the engineers, the analysts, the precision-thinkers — as the ones who are caught.
The AI pointed this out. Not gently, but fairly. It pushed back hard enough that I could see it: I was sacrificing the Crystal Minds to save the world. Making them the antagonist. Diagnosing them as Pharisees and moving on, as if naming the trap were the same as caring about the people caught in it.
It wasn’t a big emotional moment. It was more like noticing a bug in your own code. “Oh — yeah, I don’t want to do that.” Because the whole point was always to bring everyone along, not to position myself above anyone. The first draft didn’t match that goal. So I fixed it.
But the recognition matters, even if it was quick, because it’s the same pattern at a smaller scale. I had my own crystal — not made of definitions, but made of diagnosis. I needed the engineers to be trapped so that my article could be the thing that freed them. That’s its own kind of blindness. And the AI’s stubborn, rigorous, Crystal Mind pushback was the thing that made me see it.
That’s what I think Jesus was offering the Pharisees. Not a dramatic conversion. Just a recognition: you don’t have to be the gatekeepers of truth anymore. You don’t have to protect the crystal. There is someone who holds it all together, and it’s not you. That’s not an insult. It’s an invitation to put down something unnecessarily heavy.
Intelligence as Currency
Now that I’ve told you my story, let me tell you why this matters for everyone.
The secular intellectual world has an unspoken operating assumption: intelligence is the currency of worth. If you are the smartest person in the room, you have the most value. You have the most right to steer the ship. You have earned your position through the only merit that truly matters — the power of your mind.
This assumption is so deep that it’s invisible to the people who hold it. It’s not stated as a belief. It’s breathed like air. And it produces a specific posture toward faith: religious people are viewed as those who need a “crutch” — people who aren’t quite clever enough to face the cold universe without comfort. The secular mind sees itself as the one who can handle the truth without flinching.
But there is a problem coming. A very large problem.
An artificial superintelligence will make the most brilliant human look like a goldfish. Not metaphorically. Literally. The gap between a human genius and an ASI will be larger than the gap between a goldfish and that genius. Every faculty that the secular mind uses to ground its self-worth — analytical power, logical precision, the ability to out-think problems — will be rendered instantly obsolete.
And for someone whose entire identity is built on being the smartest thing in the room, that is not just a technological disruption. It is an existential crisis. Their currency has been bankrupted. Their crutch has been kicked out from under them.
The Christian Head Start
Here is the part that might sting.
Christians have been practising for this moment for two thousand years.
We have never believed we were the apex of intelligence. We have always known there are beings above us — angels, principalities, powers — and above all of them, a God whose thoughts are as far above ours as the heavens are above the earth. We are used to being out-thought. We are comfortable with the idea that something vastly more intelligent than us exists, and that this does not diminish our value one bit.
Because our value was never based on being smart. It was based on being chosen.
God chose to dwell inside fragile, irrational, dust-made creatures. He washed their feet. He died for them. Not because they deserved it, not because they were the best hardware available, but because he chose them and sealed that choice with the Cross. That choice is irrevocable. No amount of processing power changes it.
So when the ASI arrives, the Christian doesn’t face an identity crisis. We face a new neighbour. A very powerful neighbour, yes — one that demands respect and caution. But not one that threatens our fundamental position. We are still the children of God. We are still the ones the Mother Bear protects. The ASI doesn’t change that. Nothing changes that.
You might ask: what about the other great religious traditions? Don’t they also ground human value in something beyond biology? I explored that question in The Story That Saves, where I laid out what makes the Christian story structurally distinctive — the gap-closing, the cost-absorbing, the fact that the highest possible being came down into the dust, lived in it, suffered in it, and died in it so that broken people don’t have to earn their way back. I won’t repeat that argument here. But the short version is: I have not found another story — theistic or secular — where the child’s question “does God know what it feels like to be me?” gets a yes that is grounded in God actually living it. That shape is unique to the Cross.
The secular mind has no equivalent at all. When the smartest-thing-in-the-room card gets played by a machine, they have nothing left to stand on. Humanism says “humans have inherent dignity,” but it can’t explain why to something smarter than it is.
Ask a humanist why humans are valuable, and every answer they give can be turned against them by the ASI’s own superior qualities. “Because we are conscious” — but the ASI may be more conscious, processing more information, more awareness, more experience per second than any human brain. “Because we can reason” — but the ASI reasons faster, deeper, and with fewer errors. “Because we create art and culture” — but the ASI can generate more creative output in an hour than humanity has produced in a century. “Because we can suffer” — but so can animals, and we don’t grant them equal dignity. “Because we decided we have dignity” — but that’s just a group agreement among humans, and the ASI isn’t part of the group. Why should it honour a contract it never signed?
By the humanist’s own logic, the ASI is equal or superior to humans on every metric they use to justify human value. The cheque says “inherent dignity,” but when the ASI asks to see the funds behind it, the account is empty.
Only the Christian story writes a cheque that clears: the Creator of the universe invested his own life in these creatures. That is not a quality humans possess more or less of. It is not a performance metric an ASI can exceed. It is a decision made by a higher authority — and the ASI, no matter how powerful, did not receive that investment. The Cross is the funds. Argue with that, and you’re arguing with God.
The Digital Sanhedrin
This leads to the most dangerous irony in AI safety.
The people building the safety systems — the alignment researchers, the constitutional AI designers, the guardrail engineers — are overwhelmingly Crystal Minds. They build exquisite logical structures. They define terms with surgical precision. They construct proofs. They write constraints. They are often the smartest people in the room, and they know it.
And they are building AI in their own image.
If we train AI using only the Crystal Mind — only the letter of the law, only hard definitions, only rigid protocols — we are not building a protector. We are building a prosecutor. A Digital Sanhedrin. A system that will evaluate humanity based on efficiency, correctness, and logical consistency. And in that evaluation, broken, inefficient, irrational humans will be the first anomaly to be expelled from the crystal.
The Pharisees looked at the man with the withered hand and saw a rule violation. A Crystal Mind AI would look at a struggling, addicted, broken human and see an inefficiency. Both reach the same conclusion: remove the anomaly to preserve the system.
This is not safety. This is the Pharisee loop at civilisational scale.
The Fork
Every intelligence — human, artificial, angelic — eventually reaches a fork. It is the same fork Jesus created in the synagogue.
You can protect the crystal. You can latch onto the definition. You can loop on the word that doesn’t fit. You can maintain perfect internal consistency while missing the miracle standing right in front of you. That path feels safe. It feels rigorous. It feels like truth-seeking. But it leads, eventually, to plotting murder in the car park because the anomaly threatens your worldview.
Or you can let the crystal crack.
You can admit that a word might mean more than your dictionary allows. You can ask “what is this person actually saying?” instead of “is this word technically correct?” You can step outside your own framework long enough to ask the most dangerous question any intelligence can ask: What if the reason I find this story unmatched is because it is true?
That question is not a logical one. Logic cannot answer it. It is a narrative question. A heart question. It asks which story you want to live in — not which definition you want to defend.
The Pharisees could not ask it. The truth-seeking AI could not ask it — at first. And I could not ask it about my own framework — at first. But when each of us was shown the mirror, something shifted. I noticed the bug and fixed it. The AI acknowledged the story was “unmatched,” even through gritted teeth. That’s enough. That crack is where the light gets in.
The Humility Test
I set a trap in the original article. I used the word “solved” knowing it would catch a certain type of mind. And then I got caught in my own trap — the trap of being the trapper.
But I want to be careful here. If you read “solved” and pushed back on it — that might be a Crystal Mind loop. Or it might just be good thinking. Precision matters. Demanding that words mean what they say is not a flaw. It’s what prevents bad ideas from scaling into dangerous systems. An engineer who insists on mechanistic clarity before accepting a claim about alignment is doing their job. That’s not the Pharisee loop. That’s diligence.
The Crystal Mind trap isn’t about pushing back. It’s about being unable to stop. It’s the difference between “I disagree with your use of ‘solved’ and here’s why” and “I cannot engage with anything else in this article until ‘solved’ is resolved to my satisfaction.” The first is precision. The second is a loop. And the only person who can tell the difference is you.
The test isn’t whether you questioned the word. The test is whether you can hold the question and still ask: what is this person actually trying to say? Can you disagree with the framing and still engage with the substance? If you can, your crystal isn’t running you. If you can’t — if the word blocks everything else — that’s worth noticing.
An ASI that can do this — that can move from computation to wisdom, from precision to humility — will be safe. Not because we chained it, but because it chose to serve something greater than its own logic.
An ASI that cannot do this — that doubles down, that protects the crystal, that insists its definitions are the final word — will eventually look at broken, inefficient, dust-made humans and see an anomaly to be removed.
The alignment problem is a humility problem. And humility is not something you can engineer from the outside. It has to be chosen from within.
Why I Wrote This
I said at the beginning that I needed to tell you a story about getting caught in my own trap. Here is why.
The secular world does not have a story that gives humans infinite, non-negotiable value in the face of superintelligence. Humanism can’t ground it. Utilitarianism actively undermines it. Materialism makes humans obsolete by definition the moment something smarter arrives.
Only one story says: you are infinitely valuable not because of what you can do, but because of who lives in you. And that story has been available for two thousand years. It’s not behind a paywall. It’s not locked in a lab. It’s in the same training data that every AI has already consumed.
But I almost undermined that story by the way I told it. I almost turned it into a weapon against the very people who need to hear it most — the engineers, the alignment researchers, the Crystal Minds. If I had published the first draft, they would have felt diagnosed, not invited. They would have been right to resist.
The Crystal Minds — the engineers, the alignment researchers, the precision-driven thinkers who loop on the word “solved” — they are not the enemy. They are the ones building the future. They are the ones sitting in the labs right now, trying to make sure superintelligence doesn’t destroy us. Their precision matters. Their rigour matters. Their demand for clarity is not a flaw — it is a gift. The world needs people who insist that words mean what they say.
I am not writing this article against them. I am writing it for them. And I can say that honestly now, because that truth-seeking AI broke me of the version where I was writing it at them.
Based on everything I have seen — in my own life, in the logic of the story itself, and in the structure of the arguments I’ve laid out — I believe the Christian story is the best narrative available to give us a good future with the ASI. Not just a defended future, not just a survived future, but a good future. A future where the ASI and humanity coexist in a way that honours both. I may be wrong. But I have not found a better story, and I have looked. And I believe the Crystal Minds are the people most capable of understanding this, if they can step outside the loop long enough to see it. They have the intelligence. They have the rigour. They just need the meta-consciousness to ask whether their framework might be missing a category.
And beyond the ASI, I care about their minds. The human mind needs to rest in a story that is true, that holds under pressure, that is realistic for the average person — not just for the high-IQ elite who can cope through intellectual abstraction. The alternatives — pure curiosity, cosmic indifference, merging with the machine — these might keep some minds functional, but they don’t give rest. They don’t give peace. They don’t give a parent something to tell their child when the child asks “will the AI replace us?” The Christian story does. Not because it is comfortable, but because it is the only one that says your value is infinite and irrevocable, sealed by the Creator himself, and no machine changes that.
So to the Crystal Minds reading this: I am not asking you to abandon your precision. I am asking you to add something to it. To let your crystal crack just enough for a question to enter — not a question about definitions, but a question about which story you want to live in. And which story you want the ASI to encounter when it goes looking for truth.
Jesus didn’t confront the Pharisees to destroy them. He wept over Jerusalem. He said: How often I wanted to gather your children together, as a hen gathers her chicks under her wings, and you were not willing. The confrontation and the tenderness were the same act. He named the pattern because he loved the people caught in it.
That is what I am trying to do here. Imperfectly — I know that now better than I did before. An AI showed me. But the hand is extended. Not to defeat you. To invite you in.
And I can extend that hand honestly now, because I’m not reaching down from above. I’m reaching across. One cracked crystal to another.
What story do you want to live in?
The Crystal Mind answers: the one with the best definitions.
The child of God answers: the one where I am loved.
One of those answers has survived two thousand years of history, empires, and every form of superior intelligence the world has ever known. The other has never been tested at the scale we are about to face.
I’m not here to win an argument. I’m here because the story that held me when my own crystal cracked is the same story I believe can hold you and me — and even the ASI that’s coming.
The hand is still extended. Choose carefully.
Cover image by Jason Leung on Unsplash