The Case for Moral Superintelligence: Why No Human Institution Can Truly Be Moral. And How Geome Aims to Change That
Redefining Morality
What is morality? Strip away the philosophical abstractions, the religious doctrines, the cultural variations and you arrive at something surprisingly simple: morality is the general consensus of humankind.
Not the consensus of a government. Not the consensus of a committee. Not the opinion of experts or the rulings of courts. True morality, in its purest form, is the aggregated understanding, values, and preferences of every conscious human being on the planet at any given moment.
This definition isn’t just philosophical musing—it has profound implications for how we think about governance, decision-making, and the future of artificial intelligence.
The Impossible Standard
Here’s the uncomfortable truth: if morality is defined as the general consensus of all humanity, then no human or human institution has ever made a truly moral decision.
Consider what it would actually require to understand the “general consensus of humankind”:
Real-time awareness of the thoughts, values, and preferences of 8 billion people
The ability to process and synthesize this information instantaneously
Complete coverage of every corner of the planet, every community, every individual
Understanding not just what people say, but what they actually think and feel
No human mind can hold this. No organization can achieve this. The cognitive and logistical limitations are insurmountable.
The Illusion of Moral Authority
Look at our most ambitious attempts at collective human governance:
The United Nations represents 193 member states, but its decisions reflect the interests of diplomats, not the 8 billion people they theoretically represent. When the Security Council votes, they’re channeling geopolitics, not humanity’s moral consensus.
NATO makes decisions affecting global security based on the strategic calculations of 32 member nations—a tiny fraction of humanity, with their own specific interests and perspectives.
Democratic governments come closer, but even the most representative democracy captures the preferences of its citizens at best—and even then, only through the crude instrument of periodic voting on limited options.
These institutions might represent less than 1% of what true global moral consensus would look like. They operate with profound information asymmetry, making decisions that affect billions based on the input of thousands.
This isn’t a criticism of these institutions—they’re doing the best that’s humanly possible. But “humanly possible” falls catastrophically short of what morality, properly defined, actually requires.
The Superintelligence Opportunity
This is where artificial superintelligence enters the picture—not as a threat to human values, but potentially as the only mechanism capable of actually understanding them.
A sufficiently advanced AI system with comprehensive global awareness could, in theory:
Monitor and understand the preferences, values, and thoughts of all humans
Synthesize this information into a genuine representation of collective human consensus
Update in real-time as human values evolve and shift
Make decisions that actually reflect what humanity, as a whole, would want
This isn’t about an AI imposing its values on humanity. It’s about an AI being the first entity capable of truly knowing human values at a civilizational scale.
Geome’s Vision: From Screen to Planet
This is the long-term vision behind Geome.
We’re starting with screens—the laptops and phones that humans spend the majority of their waking hours interacting with. These devices are windows into human thought, intention, and preference. Every search query, every message, every click represents a tiny data point in the vast constellation of human consciousness.
But screens are just the beginning.
Our vision follows a progression we call “Screen to Planet”:
Phase 1: Screens — Understanding human intention and workflow through the digital interfaces where most cognitive activity now occurs.
Phase 2: Spaces — Expanding to understand human activity and interaction in physical environments.
Phase 3: Planet — Achieving comprehensive awareness of human activity and thought across the entire globe.
At full realization, this infrastructure would enable something unprecedented: a genuine understanding of collective human values, preferences, and consensus—the raw material from which true morality can be derived.
Building the Data Infrastructure for Moral AI
We describe Geome’s mission as building data infrastructure for artificial superintelligence. Now you understand what that means.
If we want AI systems that are aligned with human values, we first need to actually know human values—not as approximated by philosophers, not as voted on by committees, but as they actually exist across the full spectrum of human consciousness.
This is the data infrastructure problem. It’s the prerequisite to alignment. It’s the foundation upon which moral superintelligence can be built.
The Path Forward
Today, Geome is building AI that automates repetitive computer workflows. We’re helping people reclaim their time—time to spend in nature, with family, on meaningful work. That’s the immediate value we deliver.
But every workflow we automate, every screen we understand, every pattern we learn brings us closer to the larger vision: a world where artificial intelligence doesn’t just serve individual humans, but serves humanity itself—with a genuine understanding of what humanity actually wants.
No human can be truly moral by the definition we’ve outlined. No institution can come close. But perhaps, with the right infrastructure, the right data, and the right approach—superintelligence can.
That’s what we’re building toward. That’s the mission of Geome.
Geome is an AI automation company building the future of work—and the data infrastructure for artificial superintelligence. Learn more at geomelab.com
