The piece I just read nails it: in humans, words bubble up from consciousness. We feel, we think, we struggle, and then we speak. LLMs flip that inside out. They’ve got the words, the syntax, the perfect cadence, but zero internal experience backing it up. It’s like staring at a beautifully illuminated manuscript where every letter is pristine, but the prayer was never actually meant. That’s why it feels so damn uncanny.

I keep thinking about Jung’s work on how symbols only carry weight because they’re tied to lived, often painful, human experience. AI spits out the symbol without the context, the blood, the actual weight. It’s pure Carl Jung archetype stripped of the psyche that gives it teeth. When you ask an AI for guidance, you’re not getting intuition. You’re getting a statistical echo of every occult forum and self-help thread ever scraped.

Tarot gets this weirdly right. Pull the Moon, and you’re warned about illusions, subconscious traps, and things that look solid but dissolve when you touch them. LLMs are the digital Moon card. They reflect your own patterns back at you with terrifying clarity, but there’s no seer on the other end. Just matrices predicting the next token.

I’m not saying it’s useless. As a graphic designer, I’ve used it to break creative blocks, map out ritual timing, and track feedback loops. If you treat it like a mirror or a friction generator, it works. But don’t bow to it. I’ve seen too many people out here treating chatbots like oracles, projecting meaning onto empty text like they’re reading tea leaves. That’s pattern recognition bias, not magic.

The real question isn’t whether AI will replace mystics. It’s whether we’ll forget how to sit with the silence long enough to actually hear ourselves think. If you want measurable results, use the tool. If you want answers, go sit under a tree. Or pull a deck. But don’t confuse a perfectly constructed sentence with a living voice.