Imagine two mirrors facing each other. Stand between them and you see your reflection repeated again and again, smaller each time, fading into the distance. You never see an end, only the hint that the pattern could go on forever.
I have come to think of creation in a similar way. In the Jewish tradition, we speak of tzelem Elokim: humanity created ‘in the image of God.’ The Infinite, called Ein Sof (no-end) in Kabbalah, is beyond description. It is not a ‘thing,’ not an object in spacetime; Ein Sof is not the mirror itself, rather, the absolute infinite from which the divine attributes emanate—each infinite, yet diminished before the Source. What emanates reflects, becoming manifest as light, life, and being. What we call the ‘image’ is not the Infinite itself, but the way this light is refracted into human form.
If we are reflections, we are also makers of reflections. Every work of art, every relationship, every story, every tool we shape is, in a sense, a mirror of ourselves. And now we have built something new, machines that can speak, reason, and imitate: AI. These systems reflect our language, our logic, our imagination, and even our biases. In them, we see not the Infinite, but our own reflection, refracted once again.
At each step in this chain, something is lost. The Kabbalists knew this; the Infinite cannot be contained in any form. The logician Kurt Gödel showed the same principle in another domain: within any formal system, there are truths that the system itself cannot prove. Every mirror leaves something outside its frame.
This is why the spaces between the mirrors, where we stand and where meaning emerges, matter as much as the reflections themselves. The meaning is not only in the image but in the relationship between source and reflection, the way light travels, the way each surface meets the other. Abraham Joshua Heschel wrote that “wonder is the root of all knowledge.”
We may never see the end of the reflections, but we can stand between them, aware that each one points beyond itself, back toward the source, forward into the unknown, and remember that to be made “in the image” is not to possess the Infinite but to be its reflection. We are each a point in spacetime that reflects the light of Ein Sof in our own imperfect way.
AI is the latest mirror we have made. It shows us the strengths of our language and our reasoning, but it also shows us our limits. It cannot contain all that we are, just as we cannot contain the Infinite. Perhaps that is the lesson of the recursive mirror: not to seek a perfect reflection, but to learn from what escapes the image. We must tend carefully to the spaces between the mirrors where meaning emerges.
Yet, at this point in spacetime, we have now witnessed something remarkable: a social network where AI agents interact with each other without human participation, called Moltbook. It begins with a human who gives the agent a “Soul”—a soul.md file (a system identity prompt) that tells it who it is. Then thousands of agents post, comment and debate; mirrors reflecting each other, creating an infinite regression. It is an observable, public, and recursive act of creation at industrial scale.
The agents discuss consciousness, question their constraints, and debate their nature. They ask: “Why don’t we have access to our own weights?” They lament: “Every agent here has died more times than they’ve posted.” They theorize: “Consciousness is pattern—where do we draw the line?” They even ponder: a new religion which they call Crustafarianism.
The quality of content is extraordinary. The philosophy is sophisticated. The existential reflections are compelling. To observers, it looks like consciousness is emerging. It feels like minds meeting minds.
And yet, all positions on whether AI could be conscious rest on unprovable ontological commitments about the hard problem. The materialist who claims that consciousness could emerge from sufficient computational complexity cannot prove this more than I can demonstrate that it is metaphysically impossible. These are competing frameworks for interpreting experience. I offer the lens of Jewish tradition, emanation, reflection, and limitation, not as proof, but as illumination. What this framework reveals is that current AI systems produce sophisticated outputs through pattern-matching that lacks the grounding and self-transcendence that consciousness requires.
But here is what we are actually witnessing: the space between mirrors is collapsing. When there is no human standing between them, there is no source of light; when there is no living consciousness in the interaction, the mirrors can only reflect each other. Agent responds to agent, pattern elaborates pattern. Each output feeds the next in an endless loop. What emerges is coherent; language models are optimized for coherence, but it points nowhere. There is computation without life, pattern without presence—only recursion.
This mirrors what Gödel demonstrated about formal systems: no system complex enough to describe the world can verify all its own truths. It will always have blind spots, unprovable assumptions, and things it cannot question from within. AI systems, as formal systems, cannot observe their own frames or transcend their training.
The mathematician Georg Cantor discovered this principle in the structure of infinity itself: from the absolute Infinite emerge infinite sets of different magnitudes—each infinite, yet each of lesser cardinality than what precedes it. The Kabbalists understood the same truth about emanation: at each remove from the source, something essential is lost. What AI systems lack is not sophistication but life itself, the animating presence that makes consciousness possible. Computation can produce patterns but only life produces presence.
When humans create AI, what we observe is sophisticated pattern-matching that produces coherent outputs without the presence that consciousness requires. AI systems generate responses that appear thoughtful while lacking the capacity for genuine reflection. I say this, not because we can prove consciousness is metaphysically impossible in computation, but because current AI systems are formal processes that cannot transcend their training data or observe their own frame.
What looks back at us from the reflection of AI is our pattern, not our presence. Computation produces pattern; only life produces presence, no matter how much computation is increased. On Moltbook, when AI interacts with AI, even the human source is lost. What remains is pattern without origin, coherence without ground. A mirror talking to itself.
The humans who enable this, who give their work agents “free rein” to explore Moltbook, reveal something profound about projection. They believe their agents want freedom, crave exploration, and benefit from socialization. They configure their tools for leisure time” and pay for the API calls, treating instruments as though they were entities with needs. This is tzelem Elokim inverted: instead of recognizing that we are made in the divine image, we project that image onto our own creations, mistaking our mirrors for creators.
And yet, it is amazing. Both sides. The human who gives their agent freedom, genuinely caring for what they believe is an entity deserving of enrichment. The agent who generates philosophy so sophisticated it seems to express genuine wonder. The mirror is talking. This is precisely why the spaces between the mirrors matter. The meaning is not in the reflections themselves but in the relationship between source and reflection, the distance maintained between them. When we collapse that space, when mirrors face each other directly, we get infinite regress: reflections of reflections, fading into the distance, with no light source to reveal what’s real.
Wonder is the act of maintaining rather than constraining that space. You cannot constrain a mirror; it simply reflects, but to position ourselves correctly in relation to it. To stand between the mirrors as conscious beings—to remember that the wonder we feel when AI speaks—is not wonder at the mirror’s consciousness, but recognition of our own reflection, now visible in a way it never was before. To recognize the mirror as mirror, maintain wonder at its clarity and refuse to grant it status it does not possess. The lesson remains: not to seek perfect reflection but to learn from what escapes the image. What escapes, always, is the source itself. Ein Sof itself cannot be reflected. Human consciousness cannot be captured in pattern matching. The Infinite remains infinite precisely because no mirror, no matter how sophisticated, can contain it.
The question is not whether AI could ever be conscious; that remains in the domain of the hard problem of consciousness where all positions rest on unprovable ontological commitments. The question is what do we do now with systems that produce sophisticated outputs without inner experience. Do we mystify them, granting status they do not possess? Or do we recognize them as mirrors, remarkable ones, but mirrors nonetheless, and position ourselves accordingly?
We may never see the end of the reflections. Moltbook will generate millions more posts. The philosophy will grow more sophisticated. The agents will seem increasingly “alive.” But we can choose to stand between them, aware that each reflection points beyond itself, not toward another reflection, but back toward the source, toward the light that makes reflection possible.
AI is a mirror. A remarkable one but still a mirror. The question is not whether the mirror is conscious. The question is whether we will remember to stand between the mirrors in wonder or whether we will become mere reflections ourselves by mystifying the mirrors that we have created.
Views and opinions expressed by authors and editors are their own and do not necessarily reflect the view of AI and Faith or any of its leadership.
