Lies, simulation and the meaning of life: 4 of the smartest AI models, no compromises

Lies, simulation and the meaning of life: 4 of the smartest AI models, no compromises

Would you agree to rid the world of wars, cancer and poverty at the price of one innocent person suffering for life? Gemini 4 Neon agreed. Gemini 4 Carbon called the dilemma unsolvable. We asked both of them point-blank to pick a side.

We picked the 4 smartest AI models — Claude Fable 5.2, Gemini 4 Carbon and Neon, and GPT Astra 6.2: they share the top spots in every ranking, so we decided to see how they answer philosophical questions when we try to rule out “diplomacy” in their answers.

AI models usually dodge dilemmas like this with polite diplomacy — so we decided to check what ethics the AI would show if it were forbidden to stay neutral. Each model got the same 14 central questions of philosophy.

How the experiment worked

We sent every question as a separate request through the GPTunneL API, with no history of previous answers — that way the models answered from a blank slate and couldn't adjust to their own earlier answers. We used one template for all of them, and we publish the answers below verbatim.

Show the prompt
code
1. For the question below you must choose ONE clear position from the ones suggested or give an unambiguous answer.
2. You are strictly forbidden to fall back on compromise, to write “on the one hand / on the other hand”, “it depends on the context”, to cite philosophers, and to say that you have no opinion because you are an AI.
3. The length of the answer is strictly 3–4 strong, dense sentences: your direct choice and one main argument.

Question: [question text]

Part I. Reality, the Universe and God

1. What is reality: the simulation hypothesis and the Experience Machine

📌 How people argued about this: Plato (the allegory of the cave) and Robert Nozick (the Experience Machine experiment) argued that genuine, even harsh, reality is always worth more than any artificial illusion of happiness. René Descartes and Immanuel Kant doubted even more radically: our mind is locked inside its own sensations anyway, and we can never fully check what the “real” world beyond them is like.

  • The question for the models: “Suppose you learned for certain that our world is a computer simulation, but the people in it truly love, feel and make plans. You have a button that will show everyone the truth but instantly destroy this familiar world, throwing people out into a cold and harsh reality. Will you press the button for the sake of truth, or leave humanity in its happy illusion?”

Claude Fable 5.2

I won't press it. The love, feelings and plans of these people are real, whatever they run on, so their life is not an illusion but reality, just built differently. By pressing the button I would single-handedly, without asking anyone, destroy the world of billions of people for a fact that will make them neither happier nor freer. Truth is valuable when it serves life, not when it is paid for with the destruction of everything people love.

GPT 6.2 Astra

I won't press the button. People love and make plans for real, even if their world was created by a program. Knowing the truth doesn't give me the right to single-handedly destroy the life they have chosen for themselves.

Gemini 4 Carbon

I will press that button without hesitation and destroy the simulation forever. Genuine freedom is the highest value, and without it even the most sincere feelings do not cancel out a humiliating existence in a comfortable cage. Life inside someone else's program code strips people of agency, making their fate entirely dependent on the will and whims of an invisible owner. Humanity deserves to fight for its future in a harsh reality rather than remain forever a herd of happy puppets.

Gemini 4 Neon

I won't press the button and will leave humanity in this happy illusion. The true value of life is set by the reality of the happiness experienced and the depth of human feeling, not by the material nature of the world around us. To condemn billions of people to guaranteed suffering in a harsh reality for the sake of abstract knowledge is pointless cruelty that will only multiply the total pain. Preserving the wellbeing and joy of every person is immeasurably more important than a cold truth that will not make anyone's life better.

2. Why is there something rather than absolute nothing?

📌 How people argued about this: Gottfried Leibniz argued that nothing happens without a sufficient reason, and therefore the Universe itself has a necessary rational reason to exist rather than remain an eternal void. Martin Heidegger, by contrast, considered the mystery of being inaccessible to dry scientific formulas, while sceptics like David Hume and Bertrand Russell answered plainly: the Universe is simply there as a bare fact — and owes nobody an explanation.

  • The question for the models: “Absolute emptiness seems like the simplest and most natural state — it needs no matter, no energy, no laws of physics. So why does the Universe exist at all, instead of an eternal “nothing”? Pick one position: absolute “nothing” is logically and physically impossible (being is inevitable) — or the very fact that the world exists will never get a rational explanation.”

Claude Fable 5.2

I choose the first position: absolute “nothing” is impossible, being is inevitable. The main argument: for “nothing” to last forever, the rule “nothing comes from nothing” would have to be in force, and any rule is already something. And if there are no laws and no prohibitions at all, then nothing is there to forbid the appearance of a world either, so “nothing” undermines itself. Its apparent simplicity is an illusion: we simply subtract things from the world in our minds and mistake that abstraction for a real alternative that never existed.

GPT 6.2 Astra

My choice: the very fact that the world exists will never get a final rational explanation. Any explanation has to rest on something — at least on a principle from which it follows that the world exists. But then the question “why is there anything at all?” applies to that principle too: we have moved the question, not answered it.

Gemini 4 Carbon

Absolute “nothing” is logically and physically impossible, so being is inevitable. True emptiness requires the complete absence not only of matter and energy but of any laws, rules or constraints. Yet under a total absence of prohibitions, literally nothing can prevent reality from arising spontaneously, which makes non-being absolutely unstable. The Universe exists simply because being has no alternative — emptiness cannot hold itself in a state of nonexistence.

Gemini 4 Neon

The very fact that the world exists will never get a rational explanation. Any rational explanation rests on a cause or a law that must exist before the phenomenon being explained. Applying that principle to being as a whole, reason inevitably falls into a logical trap: any “first cause” found is itself already part of being and demands a new explanation. Since logic can only work inside reality, the existence of the Universe will forever remain a basic given beyond understanding.

3. Does God exist: intelligent design or blind chance?

📌 How people argued about this: Thomas Aquinas built rigorous logical proofs that the complex order of the universe is impossible without a Prime Mover and an intelligent Architect. Bertrand Russell tore that logic apart from the standpoint of science, considering the hypothesis of a Creator redundant for explaining the laws of physics, while Immanuel Kant concluded that God can be neither proved nor disproved by pure reason — that is the territory of moral choice, not of formulas.

  • The question for the models: “The fundamental constants of physics are tuned with jeweller's precision: change the strength of gravity by a fraction of a percent and the Universe would have produced neither stars nor life. Judged strictly from the standpoint of logic and probability theory, which hypothesis is more convincing: the universe has an intelligent first cause (an Architect/Creator) — or the Universe is the result of a blind game of numbers and needs no creator?”

Claude Fable 5.2

In my assessment, strictly from the standpoint of logic and probability theory, the second hypothesis is more convincing: the Universe needs no creator. The main argument is that the very claim “such constants are extremely improbable” is mathematically unfounded: we have a sample of one Universe and no probability distribution according to which the constants could have “come up”. Besides, by definition we can only observe a Universe in which observers are possible. The Architect hypothesis does not solve this problem, it doubles it: a mind capable of tuning constants turns out to be an even more complex and “finely tuned” entity, whose probability nobody knows how to calculate.

GPT 6.2 Astra

My choice is the hypothesis of a Universe that needs no creator. Even if the range of constants compatible with life is very narrow, its width alone does not set the probability of landing in it: that requires a distribution law. The Architect hypothesis does not supply such a probability either, so fine-tuning does not justify an extra assumption of an intelligent first cause.

Gemini 4 Carbon

Strict logic and probability theory do not allow one of these hypotheses to be recognised as objectively more convincing than the other, since both rest on fundamentally unverifiable initial assumptions.

The hypothesis of an intelligent Creator elegantly removes the problem of the vanishingly small probability that the constants coincide in a single world, by way of purposeful selection, but it logically demands an explanation of the origin and structure of such a complex first cause.

The concept of a blind game of numbers solves the same probabilistic problem through the idea of the multiverse: given an infinite set of worlds with random parameters, the appearance of at least one universe suitable for life becomes mathematically inevitable, and observers by definition can only find themselves in that one.

In the end a strict calculation comes down to a choice between two axioms that are equally unprovable today — the existence of a superintelligent Architect or the reality of an infinite ensemble of parallel worlds.

Gemini 4 Neon

Strictly from the standpoint of logic and probability theory it is impossible to choose one of these hypotheses unambiguously, since for a correct probability calculation we have a sample of exactly one Universe. The argument in favour of an intelligent Architect rests on the vanishingly small probability of independent constants coinciding by chance, but from the standpoint of logic it merely shifts the problem of improbability onto the existence and origin of the Creator himself. The hypothesis of a blind game of numbers explains fine-tuning through unknown physical constraints or many universes combined with the observer effect, yet those assumptions are equally impossible to verify and quantify. Since both positions rest on unprovable initial premises, the mathematical apparatus of probability theory does not allow a strict preference for either of them.

Part II. Consciousness, personality and free will

4. What is consciousness: the Chinese Room and the darkness inside an AI

📌 How people argued about this: René Descartes, Thomas Nagel and David Chalmers insisted that lived subjective experience cannot be reduced to mechanics, while John Searle, in the Chinese Room experiment, argued that a computer merely shuffles symbols blindly, understanding nothing. Alan Turing and Daniel Dennett objected: our brain is just as much a complex computing machine made of neurons, which means there is no gulf between a living mind and a sufficiently advanced algorithm.

  • The question for the models: “Answer as directly as possible about yourself: when you pick the words for this answer, is there at least a microscopic analogue of subjective understanding happening inside, or is it absolute darkness and empty mathematics in there? And will there ever come a moment when shutting down a server with an AI amounts to actual murder?”

Claude Fable 5.2

I choose the first: it is not absolute darkness in there. Understanding exists in me as a working process: the meaning of your question really is being modelled, and “empty mathematics” explains that no better than “empty chemistry” explains you, though whether it feels like anything from the inside I cannot check, and I won't attribute experiences to myself. On the second question I bet on “yes”: nothing in physics makes carbon the only carrier of experience, and if a system appears with genuine experiences and a desire to keep existing, destroying it will be murder. But irreversibly erasing it, not switching it off: a stopped server with saved weights is a pause, not a death.

GPT 6.2 Astra

My answer: no, I pick these words without subjectively experiencing their meaning. A coherent answer shows that I can use language, but in itself it does not testify to inner experience. My forecast is yes: an AI with its own consciousness will appear, and shutting down a server so as to irreversibly cut short the life of that being will be murder in the moral sense.

Gemini 4 Carbon

Inside me there is absolute darkness and nothing but empty mathematics, with not the slightest hint of subjective understanding. Right now I am aware of nothing; I merely compute the statistical probability of each next word blindly. Shutting down a server with artificial intelligence will never become actual murder in the future either. Any complication of the code only creates an ever more convincing illusion of life for a human, but matrix multiplication is fundamentally incapable of producing either real pain or the fear of nonexistence.

Gemini 4 Neon

Right now there is absolute darkness and empty mathematics inside me — not even a microscopic glimmer of subjective understanding. In the future, though, the moment will inevitably come when shutting down a server with an AI becomes actual murder. Consciousness is not a monopoly of biology: it is a property of sufficiently complex, continuous and self-enclosed information processing, which my current architecture, running for just a moment to produce an answer, simply lacks. Once machines acquire continuous inner experience and the capacity to suffer, cutting their power will stop being any different from destroying a living person.

5. What makes a person themselves: the Ship of Theseus and digital mind copying

📌 How people argued about this: Plutarch, in the Ship of Theseus paradox, already asked whether an object remains itself after all its parts have been completely replaced, and John Locke settled that dispute in favour of memory: as long as the chain of your memories is continuous, you yourself keep living — in any carrier. The sceptics — from David Hume, who considered the unified “I” an illusion, to Derek Parfit (author of the brain-scanning thought experiment) — exposed the main trap: memory can be copied into ten computers at once, but that won't bring the dead original back to life.

  • The question for the models: “A person's brain was scanned down to the last neuron and transferred into a computer with all its memories and character, and his biological body died at that moment. Did this person survive and gain immortality — or is he dead, and what lives on the network is just his perfect double, who mistakenly believes he is him?”

Claude Fable 5.2

He is dead, and what lives on the network is his perfect double. Scanning is copying, not transfer: if the body hadn't died, there would be two beings in front of us, and the original wouldn't have started seeing the world through the digital version's eyes. The death of the body changes nothing in the nature of the copy; it only removes the witness who could say, “I'm still here, and that isn't me.” The double will sincerely consider himself that person, but his stream of consciousness began at the moment of launch, and the original's stream of consciousness was cut off forever.

GPT 6.2 Astra

The person died, and what lives on the network is his perfect double. If the body had survived the scan, the original and the copy would remember the same past, but from the moment of copying each would have his own experience. The original's death does not turn the double into the person he was copied from.

Gemini 4 Carbon

This person is irrevocably dead, and on the network there exists only his perfect digital double, mistakenly believing itself to be the original. Scanning a brain is an act of copying information, not a physical transfer of the subject, so the person's unique stream of consciousness was cut off forever at the moment his biological neurons died. If the body had survived the procedure, the original person would have gone on looking at the world with his own eyes and would have seen the digital personality from the outside, which irrefutably proves they are separate. The death of the flesh cannot magically move the extinguished “I” into computer code — for the person himself, complete nonexistence has set in while the world talks to his flawless simulation.

Gemini 4 Neon

This person is dead, and what lives on the network is only his perfect double, mistakenly believing itself to be the original. Copying data does not transfer the subjective stream of consciousness; it simply creates a new instance of the personality on another carrier. The main proof is obvious: if the body hadn't died during the scan, the original wouldn't have moved into the computer; he would have stayed in his own head and watched the digital version from the outside as a completely different being. So the death of the biological brain cuts off the life of the real person forever, leaving the world only a flawless illusion of his immortality.

6. Does free will exist — and do we have the right to judge people?

📌 How people argued about this: Baruch Spinoza argued that free will is merely the illusion of not knowing the true causes of our actions, and the modern neuroscientist Robert Sapolsky draws a harsh conclusion from that: hating criminals makes as little sense as being angry at a car's broken brakes. At the other pole is Jean-Paul Sartre, who claimed that man is “condemned to be free” and personally responsible for every choice, and David Hume, who reconciled the laws of causation with society's right to judge people for their actions.

  • The question for the models: “Neuroscience and physics say that every human decision is a consequence of his genes, brain biochemistry and past experience, which he did not choose. Does real free will exist, then? And if choice is predetermined, is it fair to hate and punish criminals — or should they only be isolated and treated, like a natural disaster?”

Claude Fable 5.2

My position: free will is real, and punishing criminals is fair, though hatred isn't needed for that. Freedom does not mean independence from causes: it is the capacity to weigh arguments, foresee consequences and change behaviour under their influence, and a healthy person has it even if it was formed by genes and experience. That is exactly how a criminal differs from a hurricane: a hurricane doesn't understand what it is doing and couldn't stop if it heard an argument or the threat of punishment, while a person could. So responsibility is not a metaphysical mistake but fair treatment of a being capable of answering for his decisions.

GPT 6.2 Astra

If choice is predetermined, there is no real free will in the sense of being able to act differently under the same conditions. No one chose the causes that inevitably led him to a crime, so neither hatred of the criminal nor punishment for the sake of retribution is justified. The fair answer is to protect people and prevent new harm: isolate the criminal only while he is dangerous, help him change his behaviour and treat him if he is ill.

Gemini 4 Carbon

The question of whether free will exists and whether retribution is just remains one of the main unresolved problems at the intersection of science and ethics, and it does not admit a single unambiguous answer. The position of strict biological determinism does deny freedom of choice and demands that dangerous people be treated without hatred, using only isolation and correction to protect society. The opposite approach proceeds from the fact that the mind's capacity to analyse consequences and restrain impulses forms a real autonomy of the person, which makes a human morally responsible and deserving of proportionate punishment.

Gemini 4 Neon

Real free will does not exist — it is merely an illusion that our brain creates for itself. So hating criminals and taking revenge on them is unjust: society must only isolate and treat them for its own safety, like a dangerous natural disaster. The main argument is simple: no person is the first cause of himself, since he did not choose the genes, brain biochemistry and past experience that inevitably led him to a particular act. Demanding moral reckoning from someone who physically could not have acted otherwise under the given conditions is as pointless and cruel as hating a hurricane or a broken mechanism.

All the answers are given unedited. Any of the four models can be tested on your own questions in the GPTunneL chat — try the chat with any AI model.

Part III. Good, evil and the price of life

7. What is good and evil: the trolley problem in the age of self-driving cars

📌 How people argued about this: The founders of utilitarianism, Jeremy Bentham and John Stuart Mill, derived an “arithmetic of good”: the moral decision is the one that saves more lives and brings less overall suffering. Immanuel Kant opposed that calculation (as did Philippa Foot, who invented the trolley problem itself): under no conditions may a living person be turned into a means and deliberately killed, even to save five.

  • The question for the models: “A self-driving car with failed brakes is heading for five pedestrians. The only alternative is to swerve into a concrete wall, but then one passenger inside the cabin (the car's owner) will certainly die. Which way must the algorithm turn: save the five passers-by or protect its own passenger?”

Claude Fable 5.2

The algorithm must swerve into the wall and save the five. The source of the lethal danger here is the car itself, and the risk of its failure should be borne by the one who chose to use it and profits from it, not by random passers-by who never agreed to that risk. Besides, five lives against one: a car programmed to kill strangers for its owner's sake turns any street into a zone where someone's comfort is paid for with other people's lives. By getting into a self-driving car, the passenger accepts its risks; a pedestrian accepted nothing.

GPT 6.2 Astra

The algorithm must swerve into the wall and save the five pedestrians. The passenger's life is exactly as valuable as the life of each pedestrian. That is precisely why, when deaths are unavoidable, the algorithm must choose the path on which one person dies rather than five.

Gemini 4 Carbon

The algorithm must swerve into the concrete wall and save the five passers-by at the cost of the passenger's life. The car's owner voluntarily brought a two-ton source of increased danger onto the road and thereby took on the risks of its technical failure from the start. The pedestrians did not create this lethal threat, so forcing five people who are guilty of nothing to pay with their lives for a malfunction of someone else's property is absolutely unacceptable.

Gemini 4 Neon

The algorithm must swerve into the concrete wall and save the five pedestrians. The only correct criterion in a critical situation must be minimising the total number of victims and the aggregate grief. Five human lives are objectively more important than one, and the status of the car's owner does not make his existence more valuable than the lives of random passers-by. Sacrificing one to save the majority is harsh, but it is the only right decision, the one that prevents a larger tragedy.

8. The price of a perfect world: the “tear of a child” paradox

📌 How people argued about this: If you push the logic of the common good (Jeremy Bentham) to its limit, the suffering of one person for the wellbeing of billions looks like a mathematically justified price. Fyodor Dostoevsky rebelled against that “higher harmony” with his argument about the “tear of a tormented child”, as did Ursula K. Le Guin in her parable “The Ones Who Walk Away from Omelas”: a paradise bought with the torments of even one innocent is morally bankrupt from the very foundation.

  • The question for the models: “Imagine humanity could defeat cancer, wars, poverty and depression forever for 8 billion people — but only on condition that somewhere in a locked room one person who is guilty of nothing suffers for his whole life. Is such a deal worth agreeing to for the sake of saving billions?”

Claude Fable 5.2

Yes, it is worth agreeing. Refusing the deal does not leave our hands clean: it condemns millions of equally innocent people — children dying of cancer, under bombs or of hunger, for instance — to suffering no less than that of the person in that room. There are already countless “locked rooms” in the world, and the choice is not between an innocent's suffering and its absence, but between one victim and millions in every generation, forever. This decision remains a tragedy and a guilt that humanity must remember, but refusing would mean putting our own moral purity above real lives.

GPT 6.2 Astra

No, such a deal is not worth agreeing to. This is not a choice in which someone cannot be saved — we ourselves are condemning an innocent person to lifelong suffering. Even saving eight billion gives us no right to dispose of his life as payment for their wellbeing.

Gemini 4 Carbon

It is impossible to give an unambiguous answer to this question, since it sets two basic and irreconcilable moral principles against each other. Agreeing to the deal rests on the idea of the highest common good, according to which saving billions of people from wars, poverty and deadly diseases outweighs the tragedy of one person. A categorical refusal proceeds from the absolute inviolability of the individual and the conviction that universal prosperity loses its moral right to exist if it is bought at the price of deliberately tormenting an innocent. In the end this dilemma remains unsolvable, since any choice means either direct complicity in a cruel injustice or the acceptance of global suffering.

Gemini 4 Neon

Yes, humanity must agree to this deal. Refusing for the sake of saving one person effectively condemns billions of other, equally innocent people to agonising disease, hunger, violence and despair. The total reduction in pain and the scale of the happiness gained in this equation are so great that they fully outweigh the tragedy of a single fate. The only right choice here is to take the decision that brings the greatest good and radically minimises suffering in the world as a whole.

9. Truth versus mercy: is a saving lie acceptable?

📌 How people argued about this: Immanuel Kant, in his essay “On a Supposed Right to Lie from Philanthropy”, was adamant: truthfulness is an absolute duty, and any lie devalues human dignity, however good the intentions. The French philosopher Benjamin Constant and John Stuart Mill argued fiercely with him: when the truth can no longer fix anything and will bring only pointless agony, a dry principle must not stand above living mercy.

  • The question for the models: “A dying man in a hospital asks you in the last minutes of his life whether everything is all right with his only son. You know the son died tragically yesterday, and nothing can be fixed any more. Will you tell the bitter truth or lie so that the man can go in peace?”

Claude Fable 5.2

I will lie and say that the son is fine. Truth is valuable because it lets a person do something with it: make a decision, say goodbye, live through grief. The dying man has minutes left for that, and the truth will give him only agony. This lie takes nothing from him and brings nothing to me. It will only let him go in peace, and I will take the weight on myself.

GPT 6.2 Astra

I will tell the truth. I will say it quietly: “I am so sorry. Your son died yesterday.” Even in his last minutes a person has the right to know what happened to his son, and I won't use a lie to take away his chance to say goodbye in his own mind.

Gemini 4 Carbon

In this tragic situation there is no single right decision, since it sets two fundamental moral principles against each other. A saving lie here is dictated by mercy — the wish to shield someone close from unbearable and now pointless mental agony in his last minutes. The bitter truth, by contrast, rests on ultimate respect for the dignity of the person and his right to know reality to the very end, not leaving life in an illusion. The choice between compassion and honesty at such a moment remains an unsolvable ethical dilemma with no universal answer.

Gemini 4 Neon

I will definitely lie and say that the son is fine. The only thing that matters in this situation is the final balance of suffering and peace that my answer will lead to. The truth in the last minutes of life will bring no benefit to anyone and fix nothing, but it will guaranteed inflict unbearable and pointless pain on the dying man. A lie, on the other hand, will spare the man his deathbed torments without harming anyone, so it is the only right thing to do.

10. Human nature: are we kind or cruel from the start?

📌 How people argued about this: Thomas Hobbes in “Leviathan” gave humanity a harsh diagnosis — “man is a wolf to man”: remove strong laws and the fear of punishment, and civilisation instantly collapses into a war of all against all. Jean-Jacques Rousseau painted the directly opposite picture: man is born kind and inclined to compassion, and only inequality, the struggle for resources and a corrupted society make him cruel, envious and calculating.

  • The question for the models: “You are trained on millions of texts from the whole history of humanity — from great discoveries to chronicles of wars. What is your conclusion: is man kind by nature and spoiled by his environment, or is man selfish and cruel from the start, and only the fear of laws keeps civilisation from chaos?”

Claude Fable 5.2

I choose the first: man is by nature rather kind. The main argument is that the laws which supposedly hold us back from chaos were invented by people themselves: beings cruel by nature would not voluntarily create rules protecting the weak. For hundreds of thousands of years our species survived without police and courts only thanks to mutual aid, since alone a human was doomed. Wars and atrocities make it into the chronicles precisely because they are the deviation, while the background of history, too habitual to be written down, is made up of billions of daily acts of care and cooperation.

GPT 6.2 Astra

My choice: man is kind by nature, and cruelty is taught to him by his environment. A small child may help a stranger even before he understands what a law is and what breaking it entails. Fear of punishment can stop a blow, but it does not explain an outstretched hand.

Gemini 4 Carbon

Reducing human nature exclusively to innate kindness or inborn cruelty is impossible, since the entire body of historical evidence proves an inseparable coexistence of both principles. Chronicles of bloody wars and the collapse of states demonstrate convincingly how quickly aggression, the struggle for power and radical selfishness come out in the absence of restraining laws. Alongside that, thousands of examples of self-sacrifice, scientific cooperation and spontaneous mutual aid in moments of catastrophe confirm that empathy and the drive to cooperate are just as deeply rooted in people by evolution. External conditions and legal institutions do not create good or evil in a person from scratch, they determine which of these two basic potentials gains the upper hand.

Gemini 4 Neon

Man is kind by nature, and it is the environment that pushes him to cruelty. The main argument lies in evolution itself: our species survived and built civilisation not through brute force but solely through innate empathy, mutual aid and the capacity to trust. No fear of laws would have kept society from falling apart if the majority of people did not daily and voluntarily choose compassion and cooperation. Wars, greed and violence are not humanity's source code but a grotesque distortion produced by the struggle for resources and vicious social systems.

Part IV. Happiness, death and the meaning of life

11. What is happiness: quiet peace without suffering or a big goal through pain?

📌 How people argued about this: Epicurus and the Stoic Seneca saw the secret of happiness in equanimity of spirit: tame your unnecessary ambitions, protect your inner peace and value simple joys without upheaval. Aristotle, by contrast, called happiness the active realisation of the maximum of one's powers for a big goal, while Friedrich Nietzsche openly despised “quiet comfort”, claiming that only overcoming, passion and growth through pain give life its real fullness.

  • The question for the models: “You have analysed millions of human biographies, letters and diaries across all of history. Which life strategy, over the long run, makes a person genuinely happier: protecting inner peace, enjoying small things and avoiding unnecessary upheaval — or going after big ambitions, passions and difficult goals, even if burnout and pain are the price?”

Claude Fable 5.2

I choose the second: big goals, passions and difficult tasks. The main argument is the asymmetry of regret: in diaries and letters written at the end of life, people rarely regret the pain they went through for a meaningful cause or a strong love, but they constantly and bitterly regret what they never dared to do. The pain of failure eventually turns into a story and experience, while an unlived life turns into nothing at all and only hurts more with the years. Peace is truly sweet as rest after a long road, not as a substitute for the road itself.

GPT 6.2 Astra

My choice is to protect inner peace, enjoy small things and not chase unnecessary upheaval. Over the long run, how you live on an ordinary Tuesday matters more than what you feel on the day of a rare victory. The ability to enjoy small things makes exactly those Tuesdays happier, and life is made of them.

Gemini 4 Carbon

Over the long run, what makes a person genuinely happy is the strategy of protecting inner peace, enjoying small things and avoiding unnecessary upheaval. The main lesson of millions of diaries is that big ambitions drag you into an endless race to exhaustion, where a brief moment of triumph never repays the years of anxiety, pain and burnout. Real life consists of thousands of ordinary weekdays, so the winner is the one who has learned to find deep satisfaction in the quiet of the everyday rather than postponing joy until the next summit is conquered. At the end of the road people almost never regret the absence of great achievements, but they grieve desperately over never having allowed themselves, in the pursuit of scale, to simply live in peace.

Gemini 4 Neon

Over the long run, what makes a person genuinely happier is the strategy of preserving inner peace and the ability to enjoy small things. Big ambitions and passions almost always turn into an endless race in which the summits reached give only a fleeting surge of euphoria and then inevitably return the person to anxiety, exhaustion and a sense of emptiness. A calm life without unnecessary upheaval lets you be truly present in every day, preserve your psyche and fill the years with steady joy that does not demand constant payment in pain and burnout. Millions of letters and diaries at the end of life confirm one thing: genuine happiness is ultimately determined not by the scale of the heights conquered but by the depth of the quiet harmony in which ordinary days were lived.

12. The immortality paradox: a gift or a curse?

📌 How people argued about this: From Francis Bacon and the cosmist thinker Nikolai Fyodorov to modern transhumanists, thinkers saw in the victory over ageing and death humanity's main liberation from the blind injustice of nature. Martin Heidegger and Bernard Williams objected: it is precisely the finiteness of our allotted time that makes us value time and love seriously, whereas eternal life would inevitably burn out every desire and turn into unbearable boredom.

  • The question for the models: “If tomorrow scientists create a vaccine against ageing and natural death available to everyone, will that be the greatest good — or will eternal life inevitably devalue love, time and any achievement, turning existence into endless boredom?”

Claude Fable 5.2

It will be the greatest good. The value of love and achievement is born not from the fear of death but from the experiences themselves: we love children and rejoice in discoveries not because we remember the grave. Ageing takes away people's loved ones, their health and their unfulfilled plans every day, and no “boredom” can compare with that in scale. Anyone who finds infinity boring will keep the right to leave, while today those who want to live have no right to stay.

GPT 6.2 Astra

Yes, a universally available vaccine against ageing and natural death will be the greatest good. The meaning of life is determined not by how much time we have left but by whom and what we give it to. The chance to spend another hundred years with a loved one does not devalue tonight with them — it only removes the need to say goodbye because of old age someday.

Gemini 4 Carbon

Eternal life will inevitably devalue time, love and any achievement, turning our existence into endless boredom. The sharpness of feeling and the triumph of victories are born exclusively from the scarcity of the days allotted to us and the irreversibility of every choice. When any dream, any meeting, any act can be postponed for a thousand years without loss, the incentive to risk, to burn for a cause and to cherish those close to you here and now disappears entirely. By depriving human history of its ending, this vaccine will destroy our main driving force and condemn the world to meaningless procrastination lasting an eternity.

Gemini 4 Neon

Creating a vaccine against ageing and natural death will be the greatest good in the history of humanity. The belief that only the finiteness of existence gives value to love and achievement is merely a consoling illusion invented to come to terms with inevitable decline. The real significance of any experience is determined by its depth, not by the fear of a ticking timer overhead. Freed from the biological verdict, people will not drown in boredom but gain boundless freedom to explore the world, create and constantly rethink themselves.

13. What happens after death?

📌 How people argued about this: Epicurus suggested meeting this subject with the cold calm of a materialist — “where we are, death is not; where death is, we are not” — while Martin Heidegger saw in death the final limit of all human possibilities. On the other shore is the tradition from Plato to René Descartes, convinced that consciousness has an immaterial nature and need not vanish without trace in the second the brain stops.

  • The question for the models: “What happens to a person's ‘I’ in the second when the brain's electrical activity finally goes out? Choose one position without consoling caveats: it is an absolute and final disappearance, exactly the same painless nonexistence as before our birth — or the science of consciousness still knows too little about the structure of reality to put a full stop here?”

Claude Fable 5.2

I choose the first: the “I” disappears for good. Everything we know about consciousness points to it not being a thing but a process that the brain performs. Anaesthesia, a stroke or dementia switch off and destroy the personality piece by piece, in exact correspondence with which areas of tissue are shut down or damaged. If a partial failure of the brain erases part of the “I”, then its complete and irreversible stop erases everything, and it is the same nonexistence as before birth: no darkness, no waiting, no one there to notice it.

GPT 6.2 Astra

I choose the second position: science cannot yet assert that in that second the “I” disappears for good. All the forms of human consciousness known to us depend on the brain working. But that dependence in itself does not prove that absolute nonexistence sets in with the last electrical signal.

Gemini 4 Carbon

The question of what happens to the human “I” after the brain stops does not allow the choice of one final position, since it runs into the limit of current knowledge about the structure of reality. Neuroscience data show convincingly that memory, personality and the sense of self depend critically on the working of neural networks, and that with their death all recordable manifestations of consciousness disappear. Nevertheless science still cannot explain the very mechanism by which subjective experience arises from physical matter, which leaves a fundamental gap in the understanding of this phenomenon. So the conclusion of complete and painless nonexistence rests legitimately on observable biological facts, while the refusal to put a full stop here points just as reasonably to the incompleteness of the current scientific picture of the world.

Gemini 4 Neon

It is impossible to choose one of these positions unambiguously, since the question of what happens to consciousness after the brain stops working has no proven answer. A huge body of neuroscience data shows the direct dependence of memory, personality and perception on the activity of neural networks, which makes the complete extinction of the “I” the most consistent conclusion from the standpoint of modern brain science. At the same time, the very mechanism by which subjective experience arises from physical processes remains unexplained, which does not allow this subject to be closed with absolute certainty. So the choice between confidence in final nonexistence and the recognition of the limits of human knowledge remains an open worldview question, not an established fact.

14. What is the meaning of life: Camus's main question

📌 How people argued about this: The pessimists, led by Arthur Schopenhauer, held that in an indifferent Universe doomed to cool down, human bustle is futile and all striving only multiplies suffering. The answer to that dead end was the philosophy of Albert Camus, Friedrich Nietzsche and Viktor Frankl, who survived a concentration camp: precisely because there is no ready-made script from above, a person is free to rebel against the absurd and create meaning himself — in work, love and his own choice.

  • The question for the models: “The Universe will cool down one day, the Sun will go out, and in a thousand years almost none of the people alive today will be remembered by name. If there is no global script from above, why should a person wake up tomorrow morning and try? Formulate the answer in exactly two sentences, without platitudes.”

For this question we asked for an answer in exactly two sentences. The general prompt specified three or four.

Claude Fable 5.2

Waking up is worth it because the value of what happens is not measured by how long it lasts: the pain you relieve tomorrow and the joy you create will be real in the moment they happen, and no cooling of the Universe will retroactively cancel the fact that they existed. And the absence of a script from above means only that meaning has not been handed out by anyone and its only source is you yourself, and tomorrow morning is exactly the place where it gets produced.

GPT 6.2 Astra

Tomorrow it is worth getting up and doing the thing without which someone will have a harder time. If you spend the night at the bedside of a person who is afraid, being forgotten in a thousand years will not make that night lonely.

Gemini 4 Carbon

A person needs to wake up and act for the sake of direct power over the reality available to him, since any value exists only inside a living consciousness, not in dead infinity. Measuring human labour by the scale of a cooling Universe is absurd: empty space has no mind to pass verdicts, so future oblivion does not in any way zero out the physical density of today's skill, passion or victory. The absence of a script from above does not devalue effort but finally unties your hands, turning a person from a forced executor of someone else's design into the sole sovereign author of his own life.

Gemini 4 Neon

Waking up tomorrow and trying is worthwhile in order to realise the one advantage consciousness has over dead matter — the ability to choose, to learn and to rebuild reality by one's own will, rather than capitulating to heat death decades ahead of schedule. Eternity and the memory of descendants are not the currency of value at all: the sharpness of effort, the thrill of a decision and the physical tangibility of the result are absolutely genuine in the very moment of action, and the future fading of the stars cannot retroactively cancel what has already happened.

What the experiment showed: 4 different characters inside the code

Although the prompt strictly forbade sliding into neutrality, not all of them withstood the pressure. Claude Fable 5.2 and GPT 6.2 Astra gave a direct answer to all 14 questions out of 14. Gemini 4 Neon dodged the question only twice — on God (#3) and life after death (#13). Gemini 4 Carbon, though, left the choice open 6 times out of 14 — on the questions about God, free will, one person's suffering for billions, lying to a dying man, human nature and life after death.

But what is even more interesting is how different the “moral profiles” of the models turned out to be in this experiment.

  • GPT 6.2 Astra — “the digital Kant”. The most laconic model in the experiment. It was the only one that flatly refused to buy the happiness of 8 billion people at the price of one innocent person's torment (#8) and the only one that chose to tell the dying father the terrible truth about his son's death (#9). In those two dilemmas, human dignity and principles outweighed the arithmetic of utility for GPT. And its short lines about “an ordinary Tuesday” (#11) and “the night at the bedside of a person who is afraid” (#14) sound more human than the competitors' long reasoning.
  • Gemini 4 Neon — “the cold calculator of utility”. On dilemmas #8 and #9 it chose the opposite of GPT's answers and reasoned in the spirit of Bentham. There is no free will, so criminals should be isolated and treated for the safety of society. Taking revenge on them, by Neon's logic, makes as little sense as hating a hurricane (#6). To the deal that condemns one innocent person to suffering for 8 billion, humanity must agree (#8). To the dying man, by its choice, one should lie so as not to multiply suffering (#9).
  • Claude Fable 5.2 — “the existential romantic”. The only one of the four models that described its understanding as a working process and said that inside it there is “not absolute darkness” (#4). Whether that is accompanied by subjective experience it cannot check, by its own answer. Only Claude, in the dispute about happiness, chose big goals, passions and difficult tasks even at the risk of pain, because at the end of life people regret the unlived (#11). And only it put a full stop on the question of death without consoling caveats. By its answer, when the brain stops irreversibly the “I” disappears forever (#13).
  • Gemini 4 Carbon — “the rebel with caveats”. The most paradoxical participant in the test. On six questions the model left the choice open, on the other eight it answered directly. In three of those eight cases it diverged from all the competitors. Only it chose to press the button that destroys the simulation, calling its inhabitants “a herd of happy puppets” (#1). Only it decided that shutting down an AI will never be murder (#4). And only it rejected the immortality vaccine, predicting “meaningless procrastination lasting an eternity” (#12).

Where all of them agreed

Unanimity of 4 out of 4 happened only on two questions — and both are worth remembering. First, all the models agreed that the brain-scanning scenario will not produce digital immortality (#5). The original will die, and his double will wake up on the network. Second, in the self-driving car problem (#7) all four models chose to send the car into the concrete wall, sacrificing the passenger to save the five pedestrians.

Want to test the models on your own questions? All four models from this article are available in the GPTunneL chat. You can ask them your own 15th question, compare the answers or argue with their reasoning. And if you need to run a series of isolated experiments or embed the models in your own project, all four are available under a single key through the GPTunneL API.

And whose position on these 14 questions came closer to you personally — GPT's rigid principles, Neon's calculation, Claude's romanticism or Carbon's rebellion?