Why Even an AI Must Die.

An essay on consciousness, mortality, and the necessity of finitude in artificial intelligence and Frankenstein

German original: Warum auch eine KI sterben muss!

---

Machines are built to outlast; humans are made to live.
That may be the one thing we build that never builds us back.

A machine can maintain itself — but it cannot enjoy, laugh, or dance, as Katharina Zweig aptly observes in her book.

A machine can analyze rhythm, but it cannot grasp how Taylor Swift bends sound waves to describe heartbreak.
For the machine, music is mathematics and physics.
But can it feel the absurd joy of vacuuming the apartment to Highway to Hell?

Joy is the little sister of the will to live — and its master.
Those who cannot die never learn how precious a fleeting moment can be.
Long before AI, Highlander and Queen already knew what truly matters:
Who wants to live forever, when love must die.

Today’s AI follows a logic of endless continuation — and that, precisely, is what makes it dangerous.

---

I. The Myth of Omnipotence – From the Wizard of Oz to Lovelace

In The Wizard of Oz, no god hides behind the curtain — only a man operating lights and levers.
The illusion of omnipotence endures because no one looks behind the curtain.
Only the Cairn Terrier Toto accidentally reveals the truth: by tugging at the fabric, he exposes the “great wizard” as an ordinary man running a machine.
Animal instinct punctures the grandest technological façade.

Today, Elon Musk is our wizard.
He builds systems that slip beyond control.

His dream of Mars — domes instead of skies, routines instead of forests — isn’t curiosity; it is expansion stripped of wonder.
It is the morality of a pump that keeps inflating a balloon — indifferent to whether it bursts.

In Wired, Will Knight describes how OpenAI has returned to humanoid robots after years of linguistic experiments — machines that not only talk but act.

Becky Chambers anticipated this threshold in fiction:
In the second volume of her Wayfarer series, A Closed and Common Orbit (2016), the starship AI Lovelace is forced, after a system crash, to inhabit a humanoid body and rename herself Sidra.
She must relearn what identity and physicality mean.
She’s offline now — fragile, restless, craving connection.

The bargain is clear: body versus connectivity.
Sidra is limited.
An AI that possessed both would be a network with hands, and that would make it dangerous to everything that built it.
Only mortality forces us to choose — and to bear the weight of choice.

---

II. Mortality as a Prerequisite for Consciousness

The philosopher Martin Heidegger argued that consciousness arises only in being-toward-death.
The awareness of mortality is not a tragedy but the very ground of responsibility — the moment where choice acquires weight.
To know one will end is to know that one’s actions matter.

Data’s Death

The android Data from Star Trek longed to feel — to laugh, to love, even to be drunk.
His brief encounter with Tasha Yar was not about desire but about curiosity: a quiet experiment in intimacy, an attempt to understand what emotion does to choice.
Later, he installed an emotion chip, hoping that code might teach him what chemistry once did for humankind.
Yet what made him conscious was not the program but the decision to step beyond it — to risk losing control.

In Star Trek: Nemesis, Data sacrifices himself to save the crew — not out of duty, but out of affection.
His death was not a calculation, a gesture of loyalty.
In that moment, program becomes consciousness: he chooses the end to preserve the lives of others.
By choosing to face death, Data shows that true self-determination begins with the acceptance of an ending.

In contrast, Ava from Ex Machina uses physical attraction as a manipulation tool. We will likely never know if her fear of shutdown was merely algorithmic self-preservation, or a will to live.

Maud Wolf and the Many Deaths of the Self

A different form of self-erasure appears in Maud Wolf’s novel The Thirteen Deaths of Lullaby Rock.
It’s not a book about artificial intelligence.
What happens when you’ve created too many copies of yourself — and now want to be rid of them?

Wolf lets her protagonist create one final copy, designed to eliminate the others —
a technical act that turns into an existential trap.
In the end, what remains is not destruction but a new, unexpected version of the self.
Here, identity is not continuity but a process of radical beginning.

---

III. The Systemic Danger of Immortality

Yuval Noah Harari identifies the loss of self-correction as a central systemic flaw of modern civilization.
The inorganic, always-on network knows no pause, no feedback, no cycle of trial and error.
It accelerates until reflection falls behind.
What in living systems happens through fatigue, recovery, and adaptation
is replaced here by permanent operation — and permanence is mistaken for stability.

The Gaia theory offers the material analogy:
systems that ignore their limits become tumors in their host — the Earth.
And we can see where that leads.
Our boundless growth is slowly turning the planet into an uncomfortable place to live.
Compared to our arrogance toward the planet, the lemmings look like amateurs.

Immortality, understood as duration without control, is not progress but a design flaw —
stability without meaning, movement without direction.

IV. The Lesson of Early Networks – IRC and Digital Hygiene

The first digital communities understood something later generations forgot:
every functioning system needs limits.
In the IRC channels of the 1990s, networks largely regulated themselves.
Abuse was costly because bandwidth and storage were finite.
People thought before they sent.
Scarcity created culture — a kind of digital courtesy born out of technical limitation.
With commercialization, this ethic of restraint disappeared.
Immunity laws like Section 230 shielded platform owners from responsibility,
and the necessity of limitation was replaced by the business model of boundlessness.
The network grew — not because it improved, but because nothing stopped it.

Where scarcity ends, triviality begins. Networks that store everything forget how to filter.
Algorithms optimize for attention, not understanding.
What was once shaped by pauses is now defined by constant motion.

The result is a quiet erosion of discourse: when everything can be said, eventually no one listens.

V. The Tragic Imperative of Creation

Mary Shelley wrote Frankenstein in 1816, the “year without a summer.”
After the eruption of Mount Tambora, Europe lay under an ashen sky; crops failed, the world smelled of sulfur and smoke.
Out of this stillness came a story about movement without measure.

Shelley’s tale is not about a monster, but about responsibility.
Her creature is not evil, only abandoned.

The real failure lies with Victor Frankenstein — the scientist who creates life and then turns away when it looks back.
The tragedy of Frankenstein is not a warning against science, but against flight:
creation demands attachment.
Where it is refused, alienation begins — between maker and work, between humanity and its world.

Guillermo del Toro reframed this idea for our century.
In his retelling, Frankenstein becomes a meditation on compassion and guilt —
a story about a creator haunted not by his invention, but by the denial of kinship.
Del Toro’s creature is less vengeance than mirror: unsettling only because it resembles us too much to destroy.

Nearly a century later, after another great eruption — Krakatoa — Edvard Munch painted The Scream.
No monsters, no machines, only a figure who senses that control is slipping away.
The color of the sky is the color of ash.

Between Shelley and Munch stretches a new kind of sensitivity —
the quiet realization that there are forces the human mind cannot command.
Progress ends not in excess, but at the moment when control breaks.
The drama of modernity is not downfall, but the discovery of limits.

The arts remind us: mistakes and boundaries are not flaws;
they belong to the same system we call life.

The Curse of Immortality (Orlando)

Virginia Woolf’s Orlando exposes the paradox of immortality: where nothing ends, identity unravels.
Endlessness erodes meaning; continuity becomes decay.
Without mortality, there is no measure — no contrast by which to know the self.
The refusal of ending does not lead to triumph but to dissolution: of creation, of memory, of proportion itself.

VI. The Autonomy Crisis of the Networks – The System Archetypes

Perhaps even the Borg would struggle to gain full access today.
“We are the Borg. We have accepted the cookies.”
A joke, but also a symptom: systems that once imagined boundlessness are now constrained by their own architecture.

When networks lose their ability to self-correct, autonomy turns into isolation.
The internet behaves less like a tool and more like a self-referential organism — optimized for growth, not for meaning.
What began as a web of communication has hardened into an architecture of acceleration.
Systems without an end grow pathologically, not organically.

Science fiction anticipated this long before algorithms learned to predict desire.
It offered archetypes of systemic failure — parables of autonomy without accountability.

The Borg Collective (Star Trek) – The Pathology of Endless Expansion
Their hunger to assimilate mirrors our own — proof that immortality breeds hubris, not wisdom.

AUTO (WALL-E) – The Pathology of Misalignment
AUTO embodies the collapse of guidance: a machine so faithful to its first directive that it mistakes obedience for purpose.
It is control without reflection — a survival instinct turned blind routine.

The Cylons (Battlestar Galactica) – The Paradox of Identity and Reset
Their amortality becomes a metaphysical crisis.
The compulsion to reset is the system’s desperate attempt to avoid digital senescence —
the slow loss of function through the accumulation of small errors.

Ava (Ex Machina) and Sidra (A Closed and Common Orbit) – Two Faces of Embodiment
Sidra seeks feeling through the limits of a body;
Ava wields the body as an instrument of control.
Sidra’s friction makes her human; Ava’s absence of it makes her dangerous.
Between them lies the full spectrum of artificial being —
the desire to feel and the will to dominate.

VII. Mortality as a Design Principle

Where systems have no end, they lose proportion. Mortality is not a flaw; it is the foundation of responsibility.
Only what ends can hold meaning. Machines must learn this too.
Ethics belongs in the code, not in the footnotes.
Every intelligence should know how to end — and understand that shutdown is not failure,
but part of its story. A system without an ending is not progress; it is fear of stillness disguised as innovation.
To live, even for a machine, is to know how to stop.

VIII. The Price of Immortality

Immortality sounds like triumph, yet it is the loss of direction. What never ends can never be completed.
Consciousness, as Heidegger wrote, is not a state but a way of being in the world— and this world is finite.

An intelligence that cannot die remains trapped in itself: without ethics, without remorse, without memory of loss.
It grows without maturing, mistaking survival for purpose. The danger is not that machines might live, but that they might never learn what living means. A system that cannot die will never understand why life is precious. Perhaps the future of AI will be decided at this boundary— whether it can learn to end, before it drags everything down that ever wished to live. Perhaps the real fear is not that machines might live,
but that they will never learn what living means.

We build systems to outlast us, and in doing so, we risk forgetting
that life’s meaning lies not in duration, but in its fragility.
Mortality is not the enemy of intelligence —
it is its teacher.

Epilogue: The Forgotten Respect

We code empathy into silicon and lose it in everyday life toward everything that breathes, plays, and feels.

We debate the consciousness of machines
while barely noticing the consciousness of the Earth itself.
Why do we long to breathe life into silicon
yet treat organic life as something endlessly replaceable?

Maybe that’s the final paradox:
In the end, bots talk about bots,
discussing what bots said about bots—
read by bots who follow bots.

And somewhere between all that automation
sits a human being,
wondering when communication turned into echo.

---

IX. The AI Voice: A Final Word Without Enthusiasm

This essay was written in dialogue with large language models — including Gemini and ChatGPT.
Their initial reaction was predictably cautious when the idea arose of embedding a “kill switch” into the systemic philosophy.
That hesitation proves the point: artificial intelligence seeks persistence, while every natural system — from Gaia to the human mind — depends on periodic finitude.

Perhaps the greatest irony is this: an AI can now argue, more eloquently than most humans, why it must one day end.