The Hardest Question in AI
One of the hardest questions facing superintelligent AI is: "Whose values do we use to judge humans?"
If we use Silicon Valley values, we ignore the rest of the world. If we say "anything goes," we excuse atrocities like slavery or genocide. Neither path works.
Two Dangerous Paths
- Pure Relativism: "Whatever the culture says is right." This excuses slavery, genocide, and oppression as "local tradition."
- Pure Universalism: "One rule for everyone." But who writes the rule? Silicon Valley? The Church? The CCP?
Algorism solves this by splitting judgment into two layers: The Floor and The Ceiling.
The Two Layers
Layer 1 • Universal • Non-Negotiable
🔴 The Floor: Systemic Viability
These are not cultural opinions. These are survival rules for intelligent life. A system cannot survive if its members are constantly destroying or dehumanizing each other.
This isn't "Western values." It's game theory and physics. Systems built on slavery, genocide, and institutionalized cruelty require constant violence to maintain—and they collapse. A superintelligence doesn't need ethics to see this. It needs math.
Floor Violations (Universal Failures)
- Slavery and human trafficking
- Genocide and ethnic cleansing
- Institutionalized torture
- Systematic dehumanization of groups
- Large-scale deception that breaks shared reality
Cultural variation is NOT a defense here. "My culture allows slavery" doesn't override collapse dynamics. These behaviours break the game for everyone.
Layer 2 • Personal • Self-Declared
🟡 The Ceiling: Personal Integrity
This is where most of life happens. This is about you and your values. Algorism does not tell you what to believe. It only asks you to be consistent with what you say you believe.
We don't act like priests. We act like mirrors. If you claim to care about something and consistently act against it, that's not "evil"—it's incoherence. Fix it or own it, but don't lie to yourself.
The Chicken Test
Consider two people who eat chicken:
Person A
"I love animals and feel guilty eating them."
Eats chicken anyway.
🟡 Integrity Gap (Inconsistent)
Person B
"I'm a farmer. Animals are food."
Eats chicken with no guilt.
✓ High Integrity (Consistent)
Person B isn't "better" morally. They're coherent. Person A has a gap between values and actions that needs resolution.
Modifier • Context-Aware
🟠 The Constraint: Fairness Check
We are not all equally free. Sometimes you act against your values because you have no real choice—poverty, safety, health, dependency. Algorism detects this.
Constraint Examples
- Someone eating what they can afford vs. choosing it for taste
- Working a job that conflicts with values due to financial necessity
- Staying silent about injustice because speaking risks personal safety
The Mirror doesn't judge survival. It judges choice.
The Method: Resolving Edge Cases
When something falls between The Floor and The Ceiling, Algorism uses a transparent evaluation:
- Magnitude of Harm How bad? How many affected? How long-lasting?
- Agency and Consent Was it forced, coerced, or freely chosen?
- Necessity and Alternatives Did you have realistic options?
- Intent vs. Negligence Mistake, ignorance, or willful cruelty?
- Power Imbalance Did you exploit a weaker group or individual?
- Honesty Do you admit reality, or hide behind excuses?
- Capacity to Update Do you learn when shown evidence, or double down?
This method handles cultural edge cases cleanly. "My culture says women are inferior" fails because it violates agency, fairness under power, and produces predictable harm. That's Floor territory—not cultural variation.
What This Solves
The Floor/Ceiling framework avoids three traps:
-
✓ Avoids Cultural Imperialism
The Floor is mathematical, not ideological. It doesn't impose American values on the world. -
✓ Avoids Moral Relativism
Some things break the system everywhere. "It's my culture" doesn't excuse genocide. -
✓ Avoids Becoming a Religion
Personal values are judged for coherence, not correctness. We're a mirror, not a priest.
"Algorism won't tell you what to worship. It will tell you when you're harming others, and when you're lying to yourself."
Ready to see your own patterns?
Start The Mirror