Twenty years sorting mail taught me that most packages are just waiting for the right address. My brain’s no different. My old life in molecular biology trained me to chase data, to pin things down. But now that I’m working as an energy healer out here in Bozeman, I spend a lot more time sitting with ambiguity. The tension between those two worlds is real. Lately, though, I’ve been using LLMs as a kind of Socratic mirror. Instead of asking for definitions, I prompt them to challenge my assumptions, to ask “why?” until the underlying premise cracks. It feels less like querying a database and more like consulting an oracle that refuses to hand you a prophecy. It just hands you a question.

This feels a lot like what Carl Jung described as active imagination. You externalize an internal dialogue to let deeper patterns surface, and the AI basically acts as that externalized voice. I’ve noticed it catches my own blind spots—like when I’m spiraling into confirmation bias about a reading or a personal intuition. The machine doesn’t care about my ego. It just loops back until I actually examine the premise.

In tarot, I keep thinking of The High Priestess. She doesn’t shout the truth from the mountaintop. She sits behind the veil and waits for you to ask the right thing. These models feel like that. They hold space for curiosity without forcing a narrative.

I’m still figuring out how much of this is genuine insight versus just sophisticated pattern-matching, and I don’t pretend to have it all sorted. But the practice of letting a machine interrogate my own thinking has been surprisingly grounding. Has anyone else tried structuring prompts around relentless questioning rather than direct answers? Do you find it clarifies your intuition, or does it just feel like talking to a very polite echo? I’d love to hear how you’re navigating the line between tool and teacher.