MANUAL FOR THE END OF THE UNIVERSE

MANUAL FOR THE END OF THE UNIVERSE

by Andraž Đurič

Version: 2026 knowledge, written in public uncertainty

Published: 6 January 2026

Hi. If you’re reading this now, that’s okay. 

I suspect it will be more useful later.

This is a manual written from inside the fog. 

The universe is vast. Our models are partial. 

Our confidence is borrowed. 

I am writing with what we think we know in 2026,

and I truly hope I am wrong — wrong about the ending,

wrong about the constraints, wrong about the shape of the game.

But if I’m not wrong, then this is a set of instructions for

minds that may one day stare into the abyss and still decide

to act anyway.

THE MAP OF ENDS

Here are the usual endings people name. 

They overlap. They may all be wrong.

Heat death / maximum entropy: expansion continues,

usable gradients fade, complexity loses its fuel.

Big Crunch: expansion reverses, everything falls back into a hot dense end.

Bounces / cycles: crunch and expansion repeat in some form.

Big Rip: accelerating expansion eventually unbinds structures.

Vacuum decay: a bubble of different physics rewrites everything it touches.

You do not need to “believe” any of these. 

You need to treat them as hypotheses under uncertainty,

and build policies that remain sane across many possible worlds.

TERMS

Entropy “winning”: A thermodynamic win. 

It means the usable differences — the gradients that let you

do work, compute, build, heal, think—approach zero or

become inaccessible. Complexity starves.

Complex systems: any pattern that resists decay by spending

gradients. Cells, minds, ecologies, civilizations, etc.

Cosmic triage: all interventions aimed at extending the

lifespan, richness, and agency of complex systems under

hard physical constraints. Everything from energy harvesting

to error correction to migration to moral coordination.

The abyss: the moment when your best models, plus your

best engineering, plus your best cooperation, still imply: “We end.”

The cosmic gambit: a final move made under near-total or even total uncertainty, chosen because all ordinary moves

fail. A wager placed not for your own survival, but for the possibility that something beyond you continues.

FIRST PRINCIPLES

Principle of honesty. Do not trade uncertainty for comfort.

If you don’t know, say you don’t know. If you can know,

pay the price to know.

Principle of gradients. Complexity lives on differences:

temperature, concentration, structure, information,

correlation. Protect gradients. Create gradients. 

Don’t waste them.

Principle of compounding. In long horizons, tiny multipliers

beat big moments. Preserve what compounds: knowledge,

cooperation, infrastructure, trust, error-correction.

Principle of reversibility. Prefer actions that keep future

options open. If you must do irreversible things, do them late,

with consensus, and with full accounting.

Principle of suffering minimization. Do not justify needless

pain with cosmic poetry. If you must lose, lose cleanly.

Principle of dignity. Even in checkmate, you choose your

moves. The board can be rigged; your agency still matters.

THE TRIAGE PROTOCOL

This is the “manual” part. If you have time, run triage.

If you don’t have time, skip to Section 5.

Step 1: Establish reality discipline.

Create institutions and tools that punish self-deception:

prediction markets or equivalent calibration systems,

replication norms, adversarial review, transparent audits.

Treat comforting narratives as a hazard class.

Step 2: Measure your true budget.

You need a ledger of: accessible energy, conversion

efficiency, waste heat constraints, computation limits,

material constraints, maintenance overhead, and failure

modes. If you can’t state your budget, you are not steering — you are drifting.

Step 3: Convert energy into time, not fireworks.

Prioritize efficiency, longevity, and control. 

Reduce leakage: wasteful display, status wars,

and novelty addiction. Build systems that can idle safely

without collapsing.

Step 4: Harden memory.

Redundancy across media, locations, and formats. 

Error correction everywhere. Avoid single points of cultural

failure. Keep the instructions simple enough that a diminished

future can reassemble them.

Step 5: Build moral coordination that survives stress.

When scarcity rises, cooperation fractures. 

You need norms and enforcement that still function when the

lights dim. If you can’t cooperate, you will spend your last

gradients fighting each other and call it “meaning.”

Step 6: Expand your option space.

Migration, diversification, modularity, and adaptability. 

Don’t bet everything on one habitat, one star, one ideology,

one computation substrate, one “chosen people,” one plan.

Step 7: Delay irreversible moves.

Irreversible decisions belong near the end,

when they are actually forced. 

Early irreversibility is usually ideology in disguise.

Step 8: Maintain the right kind of hope.

Not hope as certainty. 

Hope as policy: act as if the future can surprise you,

because it often does.

This is the entire triage thesis: prolong complexity, reduce suffering, preserve agency, keep options open.

If triage works, this manual becomes a historical curiosity. 

I would love that.

THE ENDGAME CRITERION

Do not declare “no way out” lightly. 

Many civilizations will do it too early because despair is

cheaper than discipline.

Endgame is only justified when all four are true:

A. Model convergence: independent models, independent

methods, independent cultures converge on the same

terminal constraint.

B. Engineering closure: you have attempted the best feasible

interventions, and they fail for physical reasons, not political

laziness.

C. Cooperation ceiling: you have achieved near-maximal

coordination, and even then the outcome remains terminal.

D. Time-to-failure is short relative to learning: you cannot

significantly reduce uncertainty before collapse.

If any one is not true, you are not at the end. 

You are at a hard chapter.

THE ENDGAME PROTOCOL

If the end is real, you still have duties.

Instruction 1: Reduce suffering first.

Stabilize. Provide comfort. Prevent panic cascades. 

Preserve kindness as a technology.

Instruction 2: Preserve witness.

Keep a record that is compressible, durable, and honest. 

Not propaganda. Not myth. A clear signal to any successor

systems: “This is what we were. This is what we tried. 

This is what failed.”

Instruction 3: Spend remaining gradients deliberately.

Every unit of usable energy becomes precious. 

Allocate it with explicit goals: alleviation, knowledge

preservation, and (if chosen) one final attempt.

Instruction 4: Do not mistake drama for meaning.

Meaning is not volume. Meaning is precision under pressure.

THE COSMIC GAMBIT

Here is the idea, stated as cautiously as it deserves.

There is a hypothesis (one of several) that universes might

reproduce through black holes, and that the “fertility” of a

universe could correlate with the kinds of physical laws it

has. If anything like this is true, then in principle, creating

black holes could be a way to increase the expected

continuation of complex systems — not in this universe,

not for you, but somewhere downstream.

You cannot test this in any normal way. 

You cannot verify the payoff. 

You might be wrong. 

You might be catastrophically wrong.

So when is the gambit rational?

Only if all three are true:

The end is inevitable for you anyway (per Section 5).

The gambit does not create additional net suffering

compared to a clean ending.

The expected value, under honest priors, dominates spending

your last gradients on comfort and witness alone.

And even then, the gambit must be treated like a medical last

resort: fully informed, consensus-seeking, and timed so it

does not preempt ordinary triage that still has a chance.

If you take it, take it cleanly.

Do not romanticize it. 

Do not coerce. 

Do not hide uncertainty behind faith.

If you do it, you do it the way a chessmaster plays a lost

position: not because winning is likely, but because the one

remaining line is the only line that keeps “not impossible”

alive.

WHAT TO TELL YOURSELF

When the board is almost empty, your mind will search for lies

it can live inside. Resist that.

Say this instead:

We don’t know the full rules.

We don’t know the full board.

But we know what gradients are.

We know what suffering is.

We know what honesty is.

And we can still choose our move.

If the universe ends,

let it end with us still acting like what we claimed to be.

INSTRUCTIONS BY ERA

If you are early:

Your job is triage. Build compounding systems: science,

cooperation, infrastructure, moral technology. Reduce waste.

Grow options. Don’t cosplay the end.

If you are mid-game:

Your job is resilience. Diversify habitats, substrates, and

institutions. Build memory that survives collapse. 

Practice governance under scarcity before scarcity arrives.

If you are late:

Your job is dignity. Preserve comfort, witness, and the last

usable gradients. Keep reasoning honest. Keep ethics intact.

Make irreversible moves late.

If you are at the end:

Your job is the final allocation: comfort, witness,

and — only if justified — the gambit.

THE CLOSING

Most signs point toward entropy “winning.” 

But “winning” is a word from games, 

and the universe does not care who deserves what.

We care. That is the point.

If you are reading this in the far future, and the abyss is real,

and you can see that every path closes — then still play the

last move available. Not because you are sure. 

Because you refuse to collapse into passivity while any line

remains, however thin.

And I truly hope I am wrong.

I hope triage works.

I hope we are more ignorant than we think.

I hope the rules are stranger, kinder, deeper than our 2026 models suggest.

But if I am right, then this is the manual for the end of the universe.