The Human Output Hypothesis

This is a piece from Make Directory Labs, where we study operational memory, human–AI collaboration, and computational models of work. It is speculative by design — a thought experiment rather than a claim about how anything is. Our client work is next door.

If reality were a simulation, the interesting question is not who built it. It is what it is measuring.

Simulations exist to test, observe, or produce something. Weather models predict atmospheric behavior. Evolutionary algorithms search a space of solutions. Training environments expose an agent to changing conditions so that complex behavior can emerge. A simulation without an output is just an expensive screensaver.

So if this were one, it seems unlikely to exist in order to render rocks and oceans. Those look like background systems. The most complex output available inside biological reality is consciousness: a system capable of self-awareness, symbolic reasoning, moral judgment, memory, cooperation, cruelty, technology, and reflection on its own existence.

The Human Output Hypothesis is the proposal that a simulated human reality would be running to produce or measure a particular configuration of that.

DNA is the source code, not the answer

DNA is the closest natural analogue to code we have. It stores instructions, replicates, mutates, and produces organisms. So it is tempting to imagine the simulation running until the right genetic sequence appears.

That version is too simple, and the reason is familiar to anyone who has shipped software: source code is not the same as executed behavior.

Identical twins share nearly identical DNA and do not become the same consciousness. They share traits; they do not share one mind. Their experiences diverge, and with that divergence come different memories, fears, skills, and moral patterns.

DNA is the starting code. The person is the executed result. A simulator measuring human emergence could not learn what it needed by inspecting the genome — it would have to watch what the genome becomes when it runs inside a world:

DNA + environment + time + pressure + memory + choice = consciousness output

The answer is not the code. It is what the code produces.

Why the world would have to be difficult

A mind under no pressure reveals very little. To observe what consciousness actually is, the environment has to impose real conditions: scarcity, danger, mortality, imperfect information, social dependence, competing incentives, the possibility of betrayal and the possibility of love.

Which is to say a world that looks a great deal like this one — beautiful, unstable, social, uncertain, and morally ambiguous.

This does not make suffering good. From inside the system, suffering is real and reducing it remains one of the clearest signs of moral progress. But from a measurement perspective, pressure reveals structure. A painless universe might produce computation without producing courage. A frictionless one might produce preference without producing moral choice.

The output is probably not intelligence

Here is where the hypothesis stops being idle.

Intelligence alone is not a completion condition. An intelligent predator is still a predator. An intelligent manipulator is more dangerous, not less. A civilization that grows more capable without growing more restrained is not obviously a success.

So the target output cannot be intelligence. It would have to be aligned intelligence: intelligence joined to empathy, long-term thinking, self-restraint, truth-seeking, care for the vulnerable, and the ability to hold power without abusing it.

The experiment would not be can humans become smart? That has already happened. It would be can humans become wise before intelligence destroys them?

That question is not academic in a decade containing artificial intelligence, genetic engineering, synthetic biology, mass surveillance, and planetary coordination problems no existing institution is shaped to solve.

Two inheritances

Humans do not transmit only genes. They transmit culture — language, stories, values, tools, institutions, and trauma. Civilization is a second inheritance system layered on top of biology.

Genetic code builds the organism. Cultural code trains it. Together they produce historical consciousness, which means the simulation would not be purely biological; it would be informational. A family is a memory system. So is a language, a legal system, a religion, a company.

If that is right, the completion condition may not be a person at all. It may be collective: humanity becoming capable of coordinating across difference, managing existential risk, and using its own intelligence without self-destruction. Individual humans would be closer to neurons in a long historical brain than to candidates in a search.

Why it might still be running

If the simulation ends when the output appears, its continuation says something.

Humanity has achieved intelligence without alignment. Technology without wisdom. Communication without understanding. Power without restraint. Abundance in some places without justice across the whole. It has built tools capable of reshaping the planet while remaining driven by fear, status, tribalism, and short horizons.

That is not a failed experiment. It is an unfinished one.

The obvious objection

It is unfalsifiable, and the paper says so plainly. It assumes a simulation would have a purpose. It may overstate the importance of humans. It risks making suffering sound useful, which it is not — pressure revealing information about a system is not the same as pressure being justified.

The value of the framework is interpretive rather than empirical. It offers one way to hold simulation theory, genetics, consciousness, and history as parts of a single process, and it produces a question worth sitting with regardless of whether any of the premises are true:

Can life produce a consciousness mature enough to survive its own intelligence?

Everyone building powerful things is answering that question with their work, whether or not they have noticed.

The full paper

Download the white paper — sixteen sections covering the core hypothesis, DNA as substrate, consciousness as primary output, completion conditions, the moral singularity, and the objections in full.