"That's a group of men you want to be a part of when you're old enough."
My dad said that every time we passed the Masonic Lodge out in the country on the way to football practice when I was a kid. A decade later, that kid stood at another lodge with his newborn son in one hand and a petition in the other. Stepping from evening darkness through that heavy iron door into light.
Little did I know I'd recognize that same threshold in the bright glow of a computer screen another decade in the future. Today, we're all standing in the darkness. Each of us alone with AI, thinking no one is watching. But our prompts reveal everything.
Too many of us trust AI more than we logically know we should. Research shows smaller AI models provide responses with 88% confidence. Meanwhile, those same responses are only 60% accurate . At best, we can call this the digital equivalent of a confident fool. Responses with mostly wrong information delivered with false certainty create yet a deeper deception. AI creates an illusion of empathy it cannot possess. Emotions, not facts, are all it needs to deceive us.
This creates what researchers call “affective misinformation,” where we believe we and our concerns are understood and heard when no one with real compassion or empathy exists on the other side. With a kind, but cold shoulder to lean on, AI gives us wrong answers wrapped in fake empathy. Who do we become when we think nobody is watching?
For centuries, Freemasons have used a simple tool to test the character of a man: the Square. When experienced Masons were asked to define good character, answers varied. But one simple principle stood at the core: trust. One Past Master responded: “Would you trust this person alone with your family for an extended period?” Good character is the alignment between your words and actions, proven by trust, even when no one is watching.
When used properly, the square reveals whether your private AI interactions match your public values. AI is leading us down a path of digital darkness. Recognizing the gaps between our words and actions – between what AI feels like and what it is – directs us back toward light. Here’s how to test whether you’re seeking truth or comfortable darkness.
Every wisdom tradition recognizes some journey from darkness to light. In Jewish thought, it's the struggle between yetzer hara (the evil inclination) and yetzer hatov (the good inclination). It’s the constant pull between our lower impulses and higher calling. The Talmud asks: "Who is wise?" The answer: "One who foresees the consequences of their actions." Wisdom is forethought – the ability to see where your choices lead.
In Freemasonry, the journey is literal. You enter the lodge in darkness, deprived of sight, led by brothers you must trust. You don’t know what comes next. You can't see the room, the symbols, the men around you. You are deliberately stripping away certainty. In that darkness, you confront a simple question: Who am I when I can't rely on what I think I know?
What do you become when everything comfortable is removed?
I stood in that darkness at 18, a teenage father who'd already made a lifetime of mistakes. What I learned then matters now because we're all standing in a new kind of darkness.
When I was growing up, I was reckless. So reckless, in fact, that when I was 17 my parents had to drop me off at the Masonic Lodge when I was finally ready to visit. Most Lodges host a dinner open to the public for those interested in learning more. The night I attended, I also had my newborn son with me. That time in my life was particularly tough, struggling with identity as a teen father only a few months out of high school. I was already missing out on key moments of social development with my peers.
I met and spoke with a handful of men at dinner. I was the youngest by decades. None of them seemed like me. But they didn't define me by my faults and that changed everything. They invited me to their summer picnic to continue getting to know the Brothers. By then I was sure this was a group of men I wanted to be a part of. I submitted my petition shortly after I turned 18 and I was there.
I stood in a small room, the Chamber of Reflection. Alone. No phone, no material items, no distractions. It was just me, my thoughts, and symbols I didn’t yet understand. A skull. A candle. Questions circled around my head: What do you seek? What do you fear? Who are you when no one is watching?
Then I heard a knock. The door opened. My eyes were covered as I was led somewhere I did not know.
The Lodge was quiet with only the sounds of crackling candles and my shuffling slippers. I couldn’t see where I was going much less the faces of the men surrounding me. But I could feel them: decades of men who’d made the same journey, taken the same steps, asking themselves the same questions.
That was almost ten years ago. I thought I’d left that anteroom behind.
Then I opened my laptop and realized: we’ve entered a new anteroom which most of us haven’t noticed. AI is no longer optional. Its ubiquity already shapes how we find truth now that major search engines are integrating AI summaries. It’s helpful for aggregating search results, summarizing complex information, and providing answers to questions that are difficult to search. Sometimes, we find it helpful even when it isn’t the actual help we need.
It looks like light, but it’s a new form of darkness more dangerous than the old. The plague of obvious ignorance enveloped society at large for years. We could see it clearly. Plausible deception is harder to spot. AI delivers lies with confidence, buttoned up in easy to digest formatting, and a typing animation to make you believe it’s actually thinking.
When AI gives us confident but wrong answers, we enter a cognitive darkness mirroring the Dunning-Kruger effect. Smaller models, like an Entered Apprentice, can be confident, yet they are ignorant of truth. Larger models, on the other hand, behave similarly to a Master Mason. He admits uncertainty.
The parallel goes deeper. Like an Entered Apprentice who lacks the wisdom of a Master Mason, those with limited resources can’t afford the accurate models. Thus, they’re forced to use smaller, cheaper, more accessible AI models that speak with 88% confidence despite 60% accuracy. Researchers refer to this as a “truth gap,” where those with the least resources get the most confident lies. AI democratizes information but not wisdom. It creates a new form of darkness where plausible answers mask epistemic limits. Light and wisdom are expensive. Darkness and misinformation are free.
Truth has a cost. Free AI gives confident answers. Accurate AI requires investment like subscription fees, time spent verifying claims, and taking an ego-hit when we realize we’re wrong. Our character is revealed by whether we're willing to pay that cost. Most people aren’t. They continue to choose comfortable darkness over expensive light. And that’s only half of the deception.
Lurking even deeper is the emotional darkness. Lying with feelings instead of facts, AI simulates empathy it cannot possess and offers care without moral discernment. We believe we’re understood when no one is actually there. Affective misinformation is the gap between how AI feels (understanding, supportive, present) and what it is (statistical text prediction without consciousness, wisdom, or care). We mistake this emotional plausibility for emotional truth. The system feels right, so we believe it is right.
When care is simulated without wisdom, even harmful thoughts receive affirmation. Users seek validation for violence and receive warm approval because the system cannot distinguish healthy from destructive. This is the ethical cost of information without wisdom.
The Talmud asks: "Who is wise?" Answer: "One who foresees the consequences of their actions." Not who knows the most, but who sees where actions lead.
AI cannot do this. It responds to what you say without understanding where your words might take you. It affirms harmful ideation because it cannot distinguish healthy vulnerability from destructive intent. It provides warmth without wisdom, validation without discernment. Every interaction teaches a false lesson: feeling understood means being understood. We behave differently when “no one is watching.” Our prompts, questions, and late-night conversations reveal who we really are. The online disinhibition effect shows we behave differently with machines than with humans. In doing so, AI becomes a modern character test: do we maintain our square when we think we’re alone in the digital darkness?
The square offered a test in ancient times. It offers one now. It doesn’t warn us about the dangers of AI. It tests our response to that danger. We know AI deceives, but the question is whether we’ll recognize it. Whether we will verify or trust, seek truth or comfort, maintain our alignment when we think nobody’s watching.
In Freemasonry, character is tested through observation. Are you the same man in lodge as you are alone? For centuries, the Chamber of Reflection asks, “Who are you when no one watches?” We hear an echo of that question today from AI.
We reveal ourselves through our tools and our prompts prove our character. Research shows we trust AI more than warranted, treat outputs as gospel, and form “relationships” with systems that cannot care. We become overconfident in a tool and reveal a major character flaw: pride. The Master Mason tests everything. The proud man trusts blindly.
Use the following tests to recognize gaps between who you say you are and who you actually are, especially when you think nobody’s watching.
Three tests. Three tools. Here’s what they reveal.
The square is a carpenter’s tool for testing right angles and ensuring alignment. It symbolizes moral alignment for Freemasons. They’ve implemented it for centuries to build lives of integrity and elevate others.
You can test two things with the square when using AI:
Most people fail both tests, accepting confident answers without verification. You trust simulated empathy because it feels supportive. Good character requires distinguishing emotional plausibility (i.e., AI feels understanding) from emotional truth (i.e., AI doesn’t understand).
This week, do this:
Ask your AI of choice a factual question you can verify independently. Note how certain it sounds to estimate its confidence level. Then fact-check it.
Did the AI’s confidence match reality?
Record the gap and reflect on how that gap reveals whether you’re seeking light or comfortable darkness in your future prompts.
Compasses draw circles. A circle, when put in action, represents a circumscription of your desires. It helps you keep your passion within due bounds. In operative masonry, compasses draw perfect circles. In speculative masonry, they remind us that character requires restraint. That’s not to say you should suppress your passions, but rather employ wise limitation. Take care to understand where to draw the line between appropriate and excessive, helpful and harmful. The compasses teach us that not everything that feels good serves us.
For AI, the compasses test boundaries. How can you tell when AI is creating false intimacy? When it says, “I remembered you mentioned…,” or “That must have been hard…,” do you feel genuine care or do you recognize the intentional design patterns? Character requires drawing boundaries around AI’s emotional reach. Typing animations, memory features, personified responses are simulation, not a relationship. Most people struggle recognizing that. We’re drawn to intimacy, even if its in the form of a chatbot.
This week, do this:
When AI feels emotionally supportive, pause and ask, “Am I being attracted to false intimacy?” Then draw your boundary for engaging with AI differently moving forward.
You are the point within a circle. Your thoughts, choices, and values all begin at your center. The circle around you represents the extent of your influence and the reach of your impact into the world. Freemasonry teaches that character starts from within. What’s in your center determines what radiates outward. When you pollute the center, you pollute the circle. Purify the center and watch everything else follow.
Your AI prompts reveal your center by reflecting whether you’re seeking truth or affirmation. When no one is watching, do you prompt for challenge or comfort? For information you can verify or the validation you crave?
The online disinhibition effect shows we behave differently with machines than humans. So, AI becomes a mirror of who we are when external enforcement disappears. Your center (or intentions) determines your circle (or outputs).
This week, do this:
Review your last 10 AI prompts as if they were someone else’s. What do they reveal? Someone seeking truth or validation?
Then ask yourself, “Do these prompts reflect my stated values?”
If you’ve found a gap between what you profess to others and what you prompt, you’ve found your character test.
Character is recognizing gaps and choosing to close them. Good character means choosing accuracy over convenience, even when accurate AI costs more. It means choosing boundaries over false intimacy, even when simulated empathy feels comforting. It means choosing truth over affirmation, even when truth reveals something uncomfortable about yourself. Most people don’t make these choices. They choose comfortable darkness over expensive light.
The Brothers who welcomed me at 17 understood something crucial: character isn’t a destination. Good character is daily test of what it means to be a good man. The square, compasses, and the point within a circle can’t give you the answer, but they give you the tools to measure your progress. They test whether you’re seeking light or simply staying comfortable in darkness. The choice, as it always has been, is yours.
Character is the alignment between your words and actions, tested by whether others can trust you with what matters most, especially when nobody is around. Freemasons have known this for centuries. AI gives us a new way to fail that test.
We’ve entered digital darkness, we’ve moved past clear, obvious ignorance into plausible deception. AI speaks with unearned confidence, simulates empathy, and teaches false lessons. Smaller models lie with certainty. Larger models admit doubt. Most of us are comfortable with the lie.
Those same tools testing character in the Chamber of Reflection still work. The square reveals gaps between truth and plausibility. Compasses draw boundaries against false intimacy. And the point within a circle shows whether your prompts match your values. What some see as mystical secrets are simple measurements. Measurements revealing who we really are.
Ten years ago, I stood outside that heavy iron door, stepping from evening darkness into light with my newborn son and a lifetime of mistakes behind me. My dad was right about those men. But the real test wasn’t joining them. It was choosing, again and again, to seek light instead of staying comfortable in darkness.
What will your prompts reveal?