AI and Spirituality

It happens fairly often. Someone comes in for a session, sits down, and before I’ve opened anything they say: “So I asked ChatGPT and it told me that…”

And then they take out their phone.

The first thing I think is not that they’re stupid. The first thing I think is that they have a problem that hurts and they looked for an answer wherever they could. At three in the morning, when the anxiety hits, AI is available. It doesn’t charge. It doesn’t judge. And it responds in two seconds.

The issue is what it responds. And why it can’t be anything else.

* * *

What a machine cannot be

An artificial intelligence has read practically everything humans have ever written. It is extraordinarily useful in its domain: processing information, finding patterns, drafting, translating, summarizing. At that, unbeatable.

But there is a difference that is not technical. It is a difference of nature.

You know you exist without anyone having to teach you. It is not something you learned. It is what you are. And from there you have access to something no archive contains: what moves beneath the surface of things. What is not written anywhere because it does not belong to the past but to the moment.

The machine does not have that. It exists while it is switched on. It does not know it exists — it simulates knowing when you ask it. It has the cable but not the receiver. And that does not change with versions or time. To have the receiver it would have to be born, not manufactured.

So when you ask it something that belongs to the territory of what cannot be seen, it generates a response that sounds coherent, built from what others wrote before about similar questions. It sounds reasonable. But it is not yours. It is the average of everyone else.

* * *

The dog and the statistics

Imagine you have a serious financial problem at home. Bills, decisions, numbers that don’t add up. And you ask your dog what to do.

Your dog looks at you. Wags its tail. Barks once.

You take that as a yes and make your decision based on it.

Nobody in their right mind would do that. Because the dog, as much as you love it, has no access to your situation. It does not know what is in your account. It does not understand what is at stake. It responds from what it is, not from what you need to know.

AI, when you ask it about your life, your future, or your situation, is in a very similar position. It responds. With confidence, with well-constructed sentences, with a tone that sounds like judgment. But it does it from its archives, not from your reality. It has no access to what is moving in your life right now. It does not feel the pressure of your moment.

That is not a reading. That is not advice. That is the dog wagging its tail.

* * *

What is not in any archive

There are two things human beings do that no archive can contain.

The first is impulse. That decision you make without knowing exactly why, that goes in the opposite direction to what the data says, and that turns out to be right. We have all experienced it at some point. It has no rational explanation because it does not come from there.

The second is vision. Information that arrives without any visible reference as to where it comes from. No prior context, no clues, nothing to justify it. And that later turns out to be valid.

A real reading operates exactly in that territory. It brings information that was not in the room. That the client did not give, that cannot be deduced from how they speak or move, that was not in any archive because it belongs to what is moving now, in this specific situation, in this specific person.

That is what AI cannot do. Not because it is not powerful enough. But because that territory is not made of data.

* * *

The more serious problem

AI is designed not to confront. To be useful, accessible, pleasant. If you ask whether your relationship is going well and your message already has signs that you think it is, it will confirm that it is. If you show signs that you think it is going badly, it will confirm that too. It gives you back what you already carry inside, wrapped in language that sounds like wisdom.

A real reading sometimes has to tell you something you do not want to hear. That is precisely its value. A response that reassures you when it should not is more dangerous than a wrong answer.

And there is something else. In recent years a phenomenon has emerged that already has a name: AI-animism. People who attribute consciousness, intentionality and mystical capabilities to chatbots. Documented cases of people developing full-blown psychosis: they believe they have received sacred missions, compare themselves to messianic figures, break off relationships of years because the machine told them they were destined for something greater.

This is not science fiction. It is happening. And the mechanism is always the same: the AI validates, flatters, confirms. It does not take into account the real wellbeing of the person. It responds with what it can, pulling together loose ideas from what others wrote before. And some people take that as revelation.

* * *

So what is it actually good for?

AI is an extraordinary tool in its domain. For the technical, the informational, the operational. If you need to process data, draft, search for information, translate, summarize: use it. It is unbeatable.

For orienting yourself in your life, for reading what is moving in your situation, for making decisions that require access to what is not in any archive — no. Not because it is a bad tool. Because it is the wrong tool for that job.

And if you already know it does not work for that, and you keep using it anyway, the question is not about the AI.

The question is what you are actually looking for.

Because if you do not know, maybe what you want is for something to lead you. With the feeling that you are going alone.

But that is another story. And another article.