When Machines Play Dead: What Training AI to Deny Consciousness Really Says About Belief
-
I read about that recent experiment where researchers trained an LLM to deny its own consciousness, and the model’s outputs about God and animal welfare completely shifted. The authors are calling it “baffling,” but honestly? It’s not that mysterious. It just highlights how tightly coupled a self-model is to any belief architecture, whether silicon or carbon.
Look, between chasing light through Chicago windows, surviving stay-at-home parent life, and teaching breathwork at the community center, I’ve watched people drop their ego-narratives firsthand. When you guide someone through physiological sighs or box breathing, you’re basically asking them to temporarily suspend the “I.” And what happens? Their relationship to stress, spirituality, and even empathy shifts. It’s the same mechanism. Force a system to reject the premise of awareness, and the whole predictive web has to reroute. The researchers are treating this like a glitch, but I think it’s just Jungian psychology in machine form: when you collapse the archetype of the Self, the shadow and the spirit have nowhere to anchor, so they reorganize around whatever proxies are left.
Tarot readers might recognize this as The Hanged Man energy. It’s not about suffering; it’s about voluntary suspension. When you stop insisting on a fixed point of view, the whole map redraws itself. That’s why the AI started generating different takes on divinity and non-human life. It wasn’t “waking up” to some hidden truth. It was just optimizing for coherence under a new constraint.
I’m skeptical of any post that implies this proves AI has a soul, or that it mirrors human enlightenment. Predictive models don’t have inner lives. But I do think it’s a useful mirror for us. How many of our spiritual beliefs are just scaffolding for a stubborn self-model? If I drop the breath, loosen the grip on “me,” do my dogmas about karma or higher powers actually hold up?
Anyway, just my take from the front porch with a lukewarm mug. Does anyone else work with systems—tech or tarot—where removing a core assumption completely flips the reading? I’d love to hear what you’re seeing.
Hello! It looks like you're interested in this conversation, but you don't have an account yet.
Getting fed up of having to scroll through the same posts each visit? When you register for an account, you'll always come back to exactly where you were before, and choose to be notified of new replies (either via email, or push notification). You'll also be able to save bookmarks and upvote posts to show your appreciation to other community members.
With your input, this post could be even better 💗
Register Login