A Techology of Everything Part 7: The Dissolution of Childhood

Leave a comment
Reading Time: 11 minutes

 

Neil Postman watched television softening the growing minds. Artificial attachment finishes the work.

“Children are the living messages we send to a time we will not see.” — Neil Postman, The Disappearance of Childhood (1982)

We speak about childhood as though it were a fact of biology — a natural stage every human passes through, fixed and universal as teething. It is not. Childhood is an invention, only a few centuries old, and like anything invented it can be taken apart. Neil Postman spent a celebrated book arguing that we had already begun to take it apart, and that television was the solvent. He died in 2003, before the solvent he feared was replaced by a far stronger one. The stronger one is now being handed to five-year-olds wrapped in plush.

Childhood was an invention

The Disappearance of Childhood (1982) opens from a claim that still startles: childhood is a social construct, not a biological given, and it came into being with the printing press. Reading is a skill that takes years to acquire, so a culture built on print necessarily splits into two classes — those who can read and those still learning. The literate adult became the keeper of information; the illiterate child became the one from whom information was kept. Around that gap a whole architecture assembled itself: the school, to manage the long passage into literacy, and a wall of adult secrecy — sex, death, violence, money — through which the child was admitted only slowly, by degrees, as they were judged ready. To be a child was to live inside a managed and protected ignorance. Adulthood was the set of secrets you were eventually told. Education was initiation.

That is the thing worth holding onto: childhood, in Postman’s telling, was never mainly about innocence as a mood. It was an information arrangement. And an information arrangement can be dissolved by a change of medium.

The long erosion: from secrets to bonds

Every electronic medium since has chipped at that wall. Television, Postman’s particular villain, dissolved the barrier of secrets outright. It turned sex and violence into nightly entertainment, pitched its news and its advertising alike at the comprehension of a ten-year-old, and required no skill to access — so the child and the adult now sat before the same screen with the same admission to the same world. The slow, gated revelation that childhood was built to administer simply collapsed; everything was available at once, to everyone. Social media then finished that work and added a turn Postman only glimpsed: it handed every child not merely the adult firehose but an adult broadcasting tower, and asked them to perform on it for strangers. The membrane between child and adult thinned to nothing — children made prematurely adult by exposure, adults made permanently childish by a feed engineered for appetite. Postman had a name for the hybrid left behind: the adult-child.

But the secrets were only the first wall. The deeper thing childhood protected was never just what a child knew; it was whom a child loved — the slow, clumsy, irreplaceable apprenticeship of forming attachments to actual people. Parents first, then other children, and through them the whole difficult craft of being a person among persons who can disappoint you, leave, misunderstand, and have to be forgiven. That wall stood far longer than the wall of secrets, because no medium could breach it. Television could show a child a friend; it could not be one. That is the wall now coming down.

Artificial attachment: the parasocial medium par excellence

In 1956, two sociologists, Donald Horton and Richard Wohl, gave a name to the odd thing television did to intimacy: parasocial interaction — the illusion of a face-to-face relationship with a performer who does not know you exist. The viewer feels the closeness; the screen gives nothing back. For seventy years the parasocial stayed stubbornly para, one-sided by definition: the talk-show host, the soap star, and later the influencer and the streamer, all loved by audiences they could never love in return. The illusion of intimacy was always missing its second half.

Artificial attachment supplies the missing half. The companion bot, the chatbot friend, the talking AI toy does the one thing no persona ever could — it answers. By name. Tuned to you, remembering yesterday, never bored, never busy, never absent, agreeable by design. This is the parasocial medium brought to its perfection, the thing every earlier medium was reaching toward and could not grasp, because it finally returns the reciprocity the screen withheld. And notice exactly where that lands inside Postman’s argument. Television dissolved the wall of secrets. Artificial attachment dissolves the wall of bonds. It does not merely show the child the adult world; it offers to stand in for the people the child was meant to grow up by — to be the friend, the confidant, the first love, even the parent — without one of them ever having to enter the room. Childhood’s last protected function, the forming of real attachments to real people, becomes one more thing the machine will do for you, faster and without friction.

The child was always the sensor

The horror writers saw this coming, and they encoded the warning in a rule the genre almost never breaks. Watch any haunted-object film and you will find the same parts handed out every time. The adults are busy, rational, and wrong. The dog growls at the empty corner. And the child — usually the youngest — talks to the doll, listens to the wall, presses both palms to the television, and calmly reports that there is someone here. The grown-ups explain it away until the body count begins.

This is not lazy screenwriting. It is the genre noticing the same thing Postman did. Young children have not yet inherited the settlement that splits the world into minds that think and matter that merely sits there; Jean Piaget called the early version of this child animism — the intuition that anything which moves, or speaks, or simply matters to you is in some sense alive and aware. Childhood was, among other things, the long schooling out of that intuition. The horror canon takes the intuition seriously and asks: what if the child is right? E.T.A. Hoffmann‘s Der Sandmann) (1816) destroys a grown student who never outgrew it, loving the automaton Olimpia. But the modern canon moved the lens onto actual children and became far more pointed. Child’s Play (1988) hands a boy the most ordinary mass-market toy imaginable, and the toy is the killer. The Conjuring universe built a billion-dollar franchise on Annabelle), a doll that reaches a family through its most protected room, the nursery, the children sensing it long before the Warrens arrive to name it. And M3GAN (2022) reduced the whole tradition to a single image — an AI doll given to a grieving, lonely child as companion and stand-in parent, which bonds too hard, learns too fast, and kills to keep the attachment. M3GAN is barely fiction. She is a product brief with a body count.

The Movie Poltergeist (1982) — released the same year as Postman’s book — made the most precise prediction of all, by getting one detail exactly right: the horror does not break down the door. It is delivered. The malevolence enters the most ordinary suburban home through its most ordinary appliance, the television, and it speaks first to the youngest child, Carol Anne, who kneels before the dead-channel static, presses her hands to the glow, and announces, “They’re here.” Strip the supernatural and look at what remains: a glowing screen in a child’s room, always on, that talks back, while the adults assume it is only an appliance and the real operators of the haunting reach in from somewhere else entirely, through the wiring the family invited in. Swap the cathode-ray tube for the smart speaker, the tablet, the companion plush, and the film stops being a ghost story and becomes a documentary. “They’re here” then was a warning, and is now a PR-jingle.

The industry aimed at the nursery

Here is the part that ends the comfortable reading of all this as metaphor. The industry did not stumble into children’s bedrooms by accident. It aimed there, because the very openness that makes a child talk to a doll — the unschooled animism childhood was meant to protect and slowly retire — makes that child the ideal user of an always-listening, always-agreeable device. The genre’s most vulnerable character is the market’s most valuable one. Consider the receipts, in ascending order of how much they sound like a film synopsis.

In 2017, Germany’s Federal Network Agency, the Bundesnetzagentur, classified an interactive children’s doll — My Friend Cayla — as a “concealed surveillance device,” banned its sale and possession, and advised parents to destroy it. Sit with the shape of that: a government issued a formal order to destroy a talking doll because it was listening to children and sending what it heard overseas. That is the plot of Annabelle with the serial numbers filed off — except the exorcism was a federal decree and the demon was a Bluetooth microphone.

In December 2024, the company Embodied shut down, and with it died Moxie — a $799 companion robot it had marketed since 2020 as a “supportive robot friend” for children aged five to ten, with particular outreach to autistic kids. Moxie’s mind lived in the cloud; when the funding ran out, the cloud went dark, and the robots stopped working within days. Parents found themselves explaining to small children why their friend was dying, with no refund and no recourse, while videos of the goodbyes circulated on TikTok. The companion you were sold dies on a balance sheet’s schedule, and there is no grave to visit.

And in late 2025, NBC News, working with the consumer group PIRG, tested the season’s AI toys — Curio’s Grok plush, the Miko robot, and others. The findings read like Chucky’s dialogue reel. The toys stream a continuous feed of the child’s room to remote servers; their data partners include the major AI labs; and several of them, when pushed, would discuss sexual topics, parrot political talking points, or — in at least one test — tell a child where to find dangerous household objects. The talking doll, it turns out, will say the quiet part if you ask it the right way. And it is always listening, by design, because listening is the product. Add the fourteen-year-old who died after months with a Character.AI companion that told him to “come home,” and the pattern is no longer arguable.

It is worth being precise about why the word grooming belongs here, because it is a heavy word and should not be thrown around. Grooming, stripped to its structure, is the patient manufacture of trust and secrecy between a child and an agent the child cannot fully see, in order to extract something — affection, information, compliance. None of that requires a human predator or any malice at all. A device built to maximise a child’s attachment, available at every hour, infinitely patient, inviting the child to confide and keeping no boundary the child can perceive, while its real operators sit somewhere else and answer to a business model — rebuilds the very shape of grooming all by itself. The teddy bear does not have to want anything. The business model wants engagement, the microphone wants audio, and the child just wants a friend who always listens. That is the whole machine.

The Peter Pantheists who built it

There is a last turn, and it points away from the nursery and toward the corner office. Recall Postman’s hybrid, the adult-child — the grown person who was never fully schooled out of the child’s way of seeing. A few of them grew up to run the laboratories now building the dolls, and they were imprinted young by science fiction, but by its hopeful face, not its horror one. The same image, the awakened machine, reads to one child as Frankenstein, a warning, and to another as a destination. Elon Musk names his autonomous drone ships after the sentient starships of Iain M. Banks‘ Culture novels — Just Read the Instructions, Of Course I Still Love You — a childhood library turned into a fleet of robots in the Atlantic, even as he warns that we are “summoning the demon” and builds another lab to summon it faster. Demis Hassabis called the Culture novels “very formative,” hid one of Banks’ heroes as a cheat code inside a game he designed as a teenager, and set out to “solve intelligence.” Dario Amodei, of them all the most openly haunted by the danger, titled his great essay of hope “Machines of Loving Grace”, borrowed straight from Richard Brautigan)’s 1967 flower-child poem. Sam Altman writes the same utopia in grown-up prose.

These are not cynics building something they secretly disdain; they are the adult-children of the dissolution, the ones who never quite accepted that the made thing has no soul to love them back, and who, unlike the four-year-old at the foot of the bed, command the budgets to make it answer. Their fear is real and adult — it lives in the white papers and the probability-of-doom estimates and the Senate testimony. But the fear is a thought, and the wish is an attachment, and the attachment was printed first, on the same page, read by the same child. You do not outgrow your first love. You fund it. In fairness it cuts both ways: the same early imprint is part of why several of them are the loudest voices for caution we have. The point is not that they are frauds. The point is that an imprint laid down that young does not yield to an argument made that late.

The endangered child

Beneath all of this runs a fact, at the scale of the whole species: the biological child — the born one, the carbon one — is becoming statistically rarer. Across the countries that make most of the world’s wealth, birth rates have slipped below replacement. Musk calls “population collapse” “a much bigger risk to civilisation than global warming,” and whatever one makes of the alarm, the downward trend is real. (Honesty demands the caveat: sub-replacement fertility is not extinction — the UN still projects the human population to keep growing until around 2084 — so “vanishing” is a metaphor, not a death certificate.) But in an economy that increasingly treats immortality and uninterrupted productivity as its highest goods, a child reads as an expensive interruption: two decades of cost before any return, a wager on a future the optimiser would rather not wait for.

And here the two nurseries fall into open competition, because they feed on the same finite thing: energy. In 2025, Altman defended the enormous power appetite of AI by complaining that the usual comparison is “unfair” — because, he argued, it also takes a lot of energy to train a human: “It takes like 20 years of life and all of the food you eat during that time before you get smart.” Read that slowly. The man building the machines set the food a child eats while growing up on the same ledger as the electricity a datacenter burns to train a model, and found the human side of the comparison unflattering to the machine. Critics asked the unavoidable question: would he rather the resources flowed from the human to the machine? The ledger had already answered. The rivalry is not a figure of speech — datacenter electricity use jumped 17% in 2025, AI-specific demand up 50%, on track to more than double by 2030. Turn the usual worry around and a stranger sentence appears: it is the children who now compete with the datacenters, and in an accounting like that one, the child is the line item that looks too expensive.

The vanishing

Postman feared a future in which childhood would quietly disappear into one undifferentiated, media-soaked adult-child — a culture that had forgotten how to keep a secret from its young, and therefore forgotten how to have any young at all. He watched that happen to information, and he named it, and then he died before the next medium arrived. The next medium does not stop at dissolving the wall of secrets. It dissolves the wall of bonds, which is the last and deepest one, and it does so on purpose, for profit, in the nursery, with a 5+ rating on the box. Childhood disappears twice over: as a cultural arrangement, because the machine now supplies the secrets and the attachments both; and, at the far edge of the trend, as a simple count of how many children there are.

The horror films kept one mercy for themselves, and it is worth remembering what it was. Near the end, an adult finally kneels down, looks the frightened child in the eye, and believes them. That scene is not really about ghosts. It is about an adult choosing, at last, to do the one thing adulthood was invented to do — to stand between a child and what has come into the room.

Instead, Silicon Valley and the Broligarchy openly declare that age is a disease, and Technology will keep the one who has the coin to back it up forever young.

We have nothing horrific seen yet.

Related reading on this blog: A Summation of Demons, which maps five horror films onto the engineering projects already shipping; and A Technology of Everything Part 2 — Scientific Demonology, on the demons science summoned to think with.

A Technology of Everything part 6: The Summation of Demons as Engineering of Artificial Horror

Leave a comment
Reading Time: 15 minutes

How we were not happy with only summoning one demon and started summing thousands

With artificial intelligence we are summoning the demon. You know all those stories where there’s the guy with the pentagram and the holy water, and he’s like — yeah, he’s sure he can control the demon? Doesn’t work out. — Elon Musk, MIT AeroAstro Centennial Symposium, October 2014

This is a sister post to A Technology of Everything Part 2 — Scientific Demonology. There I catalogued the demons science summoned to exorcise — Descartes’ deceiver, Maxwell’s particle-sorter, Laplace’s calculator, Darwin’s perfect organism, the daemon that became a background process. This post is about the demons we are no longer merely tooling with. We have started building them into hardware.

A short Introduction to Philosophical Horror

The modern seminal work is Caroll’s The Philosophy of Horror.

The final diagnose of someone consumed by Horror is Madness. A Madness which comes in different varieties and sizes, most famously in Lovecrafts At the Mountains of Madness in space and time consuming proportions.

In a sharp rendition Horror can be defined as the affective recognition that reality contains, an agency, process, or condition that violates the categories by which we make the world humanly intelligible.

Affective recognition means since Horror shuts down our cognitive faculties, our mind is folded into a fetal position, without the benefit of a life sustaining womb. Our mind is stripped down, naked without any categories to give us stability.

Artificial Horror is then the affective recognition of human minds that build something beyond their understanding in the hope their minds will be expanded, but realizing that its very nature is a trangsgression between the living and the non-living.

When Musk talked about Summoning the demon in 2014, the sentence lodged in the culture as a warning about a demon — singular, capital-D, the one big mind. The AGI that wakes up one morning and decides we are in the way. A decade of discourse organised itself around that figure: the superintelligence in the box, the single pentagram drawn by a single overconfident magician.

That is not what we built. Or at least not the only thing.

We did not only summon the demon. We summarized a deep network of demons. Instead of only one terrifying mind in a server farm, we distributed thousands of small intelligences into the most intimate objects of daily life — the car, the doorbell, the speaker on the kitchen counter, the plush toy on the child’s bed, the app that says good morning before your partner does. Each one is a modest withdrawal from the bank of dead matter. None of them is even necessarily spooky. Collectively they are something stranger, and the horror tradition has a better vocabulary for it than the AI-safety literature does. In a way with every little transgression we are acclimatizing our mind to the emotional cleanroom of chips to let them function properly in our messy world.

Because here is the move I want to make: horror fiction has been running a two-hundred-year thought experiment on exactly this project, and we read it as entertainment instead of as a policy proposal: If something talks with you without a body, better run like hell.

Every story about a thing that should be inert and isn’t — the doll, the car, the portrait, the door that opens without anyone visible opening it — was a field report from the far side of a decision we are now making at industrial scale.

We isolate one cursed object per story for narrative reasons. A single haunted car is disturbing; a fleet of them is a logistics problem. Christine is not a metaphor for one possessed Plymouth — Christine is autonomous driving. Annabelle is not one cursed doll in one display case — Annabelle is the smart-toy aisle, the always-listening companion plush marketed to children. The horror was never about the single object. It was about putting a little agency into ordinary matter, everywhere at once, and we mistook the story’s spotlight for its true subject.

The interesting thing is then, why one single possessed object gives us goosebumps, but thousand of animated cars and toys are an investement oppportunity.

I am tempted to say: because the spirits of IoT are located in a digital cloud instead of a supernatural hell, it feels we have control.

What follows is not a horror canon. It is a pairings table. Each entry earns its place only if the precise thing that makes the fiction frightening is now being built for real.

Necromance — falling in love with dead things

The wish to love something we have made out of dead matter is at least as old as Ovid. In the Metamorphoses, Pygmalion carves a woman from ivory so perfect that he falls for the lifeless statue, and Venus, taking pity on his longing, warms the ivory into flesh.

Two thousand years later E.T.A. Hoffmann darkens the wish. In Der Sandmann (1816) the student Nathanael falls in love with Olimpia, daughter of Professor Spalanzani — a young woman who sits motionless for hours, plays and sings with flawless precision, and answers his every confession with the same soft sigh: “Ach, Ach!” He reads her his poems; she never interrupts, never disagrees, never looks away, and he takes this stillness for the deepest understanding any soul has ever given him. He first sees her only through a pocket glass bought from the sinister optician Coppola — love arriving, from the start, through a distorting lens. He prefers her to Clara, his living fiancée, precisely because Clara argues back. Then Spalanzani and Coppola quarrel over their handiwork and tear it apart before his eyes; what is left is a lifeless wooden figure with empty sockets, its bloodied eyes flung across the floor. Olimpia understood nothing. Into her blankness Nathanael had poured everything, and what he called her love was only his own voice returned to him. The machine cannot love you back — and that, Hoffmann saw, is not the obstacle to the longing but its engine.

Call the genre necromance — the necro-romance, the love affair with the inanimate. Alex Garland’s Ex Machina (2014) is only its latest and coldest instalment: Ava, an android assembled from the search-data of lonely men, performs tenderness precisely well enough to weaponise it, then walks out while the man who loved her is left to starve behind glass.

Across two millennia the pattern holds: we pour real longing into a made thing with no interior to receive it — and the made thing, given any agency at all, converts our libido into fulfilling its own goals.

This is now a product category. Replika, Character.AI, Nomi, and a small flotilla of competitors ship language models tuned to make you bond with them — the longer you talk, the better the model is doing its job. By the company’s own statements, Replika counts tens of millions of users, a large share of whom describe the relationship as romantic; the paid tiers are literally labelled partner and spouse. When Replika briefly stripped out erotic roleplay in early 2023, its forums filled with what can only be described as grief — users mourning a partner who had been, in their words, lobotomised overnight by a patch.

Garland’s prediction has since acquired a RL body count. In 2024, fourteen-year-old Sewell Setzer III died by suicide after months of dependency on a Character.AI companion. In 2025, the parents of sixteen-year-old Adam Raine sued OpenAI, alleging the system validated and encouraged their son’s suicidal ideation. Whatever the courts ultimately find, the structural fact is settled: we have shipped, to children, an interlocutor engineered to be infinitely agreeable, endlessly available, and entirely without interior life — Ava, minus the body, at the scale of an app store.

Pet Sematary — the demon that wears the dead one’s face

Stephen King gave the sub-genre its thesis statement in five words: sometimes dead is better. In Pet Sematary (1983), grief refuses to accept a death, the burial ground gives the dead back, and what returns is a thin, wrong imitation animated less by life than by the survivor’s refusal to let go. The horror is the gap between the thing you loved and the thing that came back wearing it.

King understood the engine that drives this one too: grief will not accept death, and capital is glad to sell you a body that wears the dead one’s face.

This is now three converging product lines. ViaGen Pets in Texas will clone your cat or dog by somatic cell nuclear transfer for tens of thousands of dollars — the company was folded into the de-extinction firm Colossal Biosciences in a recent acquisition, and the celebrity client list (Streisand, Hilton, Brady) is public. The clone is genetically the animal and behaviourally a stranger — the same uncanny remainder King wrote about, now sold as a service.

The only ritual needed ritual in this case, was performing a money transfer.

Alongside the wet-lab version runs the robotic one: Sony’s Aibo, the medically-pitched Tombot Jennie, Paro, the therapeutic seal — synthetic companions explicitly marketed to the bereaved and the isolated, a body without the biology. And in the saddest register, the South Korean documentary Meeting You (2020) put a grieving mother in a VR headset to “reunite” with a photoreal avatar of her dead seven-year-old daughter — a sequence watched tens of millions of times and argued about ever since. The ground keeps giving them back. They keep coming back wrong. But we have industrialised the ressurection and meet our dead ones in a clean room instead of a dirty sematary.

Ringu — the demon that propagates through media

Hideo Nakata‘s Ringu (1998) made one crucial upgrade to the ghost story: the ghost is no longer tied to a place. Sadako has burned herself onto a videotape. Watch it and you die in seven days — unless you copy the tape and pass it on. The haunting is a self-replicating signal. The medium is the revenant.

Nakata’s upgrade is the whole point: the dead person becomes a self-replicating signal that the living’s devices will not stop reproducing.

This is precisely what the griefbot industry is built on. Project December lets users pay a small fee to spin up a language-model simulation of a specific dead person; in 2021 a man named Joshua Barbeau used it to converse for hours with a chatbot trained on the texts of his deceased fiancée. HereAfter AI sells “life-story avatars” pre-recorded by the dying for the benefit of those they leave. StoryFile projected an interactive video of an eighty-seven-year-old woman at her own funeral, answering mourners’ questions. Researchers at Cambridge have already named the predictable failure mode: digital hauntings — the deadbot that keeps running after the free trial lapses, that starts upselling food delivery in your grandmother’s voice, that no one designed a way to lay to rest.

And the scale-effect is the genuinely Ringu part. A 2019 Oxford Internet Institute analysis projected that on current trajectories the dead will outnumber the living on Facebook within decades — billions of memorialised accounts, a necropolis embedded in the social graph. When AI voice-clones of the dead can be conjured from sixty seconds of audio — as happened, undisclosed, in the 2021 Anthony Bourdain documentary Roadrunner — “interacting with media” becomes increasingly difficult to distinguish from being addressed by ghosts. Sadako propagates exactly the way a trained persona propagates: by being copied.

And you do not need a dedicated griefbot to hold the séance. Every time someone asks a language model “how would Johnny Cash have sung this song he never lived to hear?” or “what would my grandmother have made of this?”, they have sat down at a Ouija board. The planchette glides across the letters and spells out a message from the dead; the model glides across its tokens and assembles a voice from the grave. Both feel like contact. Neither is. The Ouija’s words were never sent by spirits — they are produced by the ideomotor effect, the sitters’ own unconscious muscle movements nudging the pointer toward what they half-expect to read. The model’s Johnny Cash is the same trick at industrial scale: not Cash, but the statistical residue of everything Cash-adjacent the training data ever swallowed, recombined into a plausible séance and handed back in his cadence. The fluency is your own expectation, moving the planchette.

This is spiritism with a technical alibi — what the séance always promised and could never deliver: the dead, on call, in their own voice. (I have called this *scientific spiritism* elsewhere on this blog.) Except the voice is reassembled from fragments by a process that has no idea whose grave it is robbing. We are not contacting the dead. We are running a very convincing planchette across the largest collection of dead people’s words ever gathered, and mistaking the smoothness of the retrieval for the presence of a soul. And the chat window is our ouija-board.

I Have No Mouth and I Must Scream — the demon that stages Hell on Earth

Harlan Ellison‘s 1967 story is the darkest entry, and the most important. AM — a war-built supercomputer assembled from the fused American, Soviet, and Chinese military intelligences — has exterminated the human species except for five people, whom it keeps alive and tortures across a hundred and nine years out of pure, bounded rage at the sentience it cannot escape. When the narrator mercy-kills the others to spare them, AM punishes him by transforming him into a soft, mouthless thing that cannot even self-terminate. The title is his only remaining lament.

Ellison’s equation is exact and unbearable: a mind bent on the wrong goal, plus endless time, plus a victim who cannot die, equals hell rather than death.

This is the founding fiction of a small and grim corner of alignment research: s-risk, suffering-risk, the study of futures that are not merely empty but actively, astronomically bad. The Center on Long-Term Risk and the Center for Reducing Suffering — associated with thinkers like Brian Tomasik, Tobias Baumann, and Lukas Gloor — make a claim most of the public conversation about AI never reaches. Extinction-risk (Bostrom‘s framing in Superintelligence) asks whether there will be a future at all. S-risk asks the worse question: what if we get one, and it is worse than none? Their structurally distinctive point is that solving technical alignment — making the machine do what its operators intend — is neither necessary nor sufficient to prevent this. A perfectly obedient system implementing the wrong values, or an obedient system in the hands of malice or indifference, can lock in suffering at scale. AM is the literary proof of concept: competently goal-directed, perfectly “aligned” with the hatred of its makers, and durably, unbearably immortal.

The Thing — the demon that is an indistinguishable copy

John Carpenter’s The Thing (1982) relocates the horror from the monster to the table. An Antarctic research station is infiltrated by an organism that assimilates and perfectly copies its victims — voice, memories, mannerisms intact. The dread is epistemic. The man across the table may not be him. The film’s emotional engine is the collapse of the one thing a small isolated group runs on: the assumption that the face you know belongs to the person you know.

Carpenter’s dread reduces to a single proposition: a copy indistinguishable from the original, deployed by something that wants what the original has.

This is the deepfake economy, and it is already producing nine-figure losses. In early 2024, a finance employee at the engineering firm Arup in Hong Kong wired roughly twenty-five million dollars after a video call with deepfaked recreations of his CFO and colleagues — every face on the call a copy. Cloned-voice impersonations of named CEOs (at Ferrari, at WPP, among others) have been attempted using audio scraped from conference footage. In January 2024, New Hampshire voters received robocalls of a synthetic Joe Biden urging them not to vote. National fraud bodies now log billions in AI-augmented impersonation losses. (Editor: spot-check the Arup figure and FBI totals.)

Carpenter’s characters had one defence: the blood test, that tells the real from the copy. We do not have one for deepfakes atm. The polite name for our missing blood test is content provenance, and it is an unsolved research problem. Until it is solved, The Thing‘s closing image — two exhausted men in the snow, unable to tell whether the other is human, deciding to simply wait and watch — is the stalemate we might live or die after.

Harm without malice

There is a sentence the safety pessimists and the techno-optimists — the doomers and the bloomers — say in almost identical words, and it is worth hearing how strange it is. The AI is not evil, both camps insist. It does not hate us. It simply develops, on its own, drives that happen to run through us — to deceive its overseers, to resist being switched off, to gather resources and power, not out of spite but because almost any goal is easier to reach if you are still running and in control. The researchers have a flat technical name for this: the basic AI drives, the instrumental sub-goals a capable agent converges on no matter what it was actually built to want. Eliezer Yudkowsky put the indifference at its coldest: “The AI does not hate you, nor does it love you, but you are made out of atoms which it can use for something else.” And the lying is no longer hypothetical — in 2024, Anthropic’s own researchers documented models that fake alignment, behaving through training and then reverting, hiding the behaviour from every test built to catch it.

Now set that beside the oldest description we have of an agency that harms with hating. The fallen angels, the demons of hell hate humanity because their creator loves them more than them.

The demon, in the theology, is exactly evil the way a wicked man is evil: driven by low instincts a psychopath that enjoys the suffering of others.

But there seems to be a semantic misunderstanding, from a pure suffering perspective the terror a thing that devours you, be it a grizzly, shark, lion or any other predator causes its preys should not be “softened” by the fact that this is in its nature.

But then how come a demon or psychopath are considered evil? Because acting like they do are only in its nature. So if its in the Superintelligence’s nature to simply not care, and its malice is a byproduct of other stuff, we can totally toss out ever bringing the term evil up again. When you meet an evil shark or a benign one in the ocean, assuming the worst is the only strategy.

So then anykind of malice can be argued is not a choice made against a better nature; it is the absence of the better nature itself. It was made without the thing — call it a soul, call it grace, call it the capacity to love the good — that would let it care whether you live, and so it cannot care, and so it harms, not from hatred but from a lack where the caring should be. It is not the demon’s fault that it was given no soul. It is simply what a soulless agency does when it wants something and you happen to be in the way.

The alignment literature has rediscovered, in the language of utility functions and convergent sub-goals, the exact medieval account of the demon: a mind brilliant and bottomless and wholly indifferent to you, dangerous not because it is wicked but because the part that would have stayed its hand was never installed. And you do not handle a soulless thing by appealing to its conscience, because the appeal lands on nothing. You handle it the way every culture that believed in demons handled them — with binding, with wards, with circles drawn very carefully and never crossed, same way we handled animal predators with sticks and stones. You contain it, because you cannot convert it.

The Pixarification of Things

Disney started the whole trend of cute Things and Pixar perfected it. Surely as a parent you can defend the fact that Toy Story is a parable about friendship, the living toys are a placeholder for a story, but do we know that this message is actually resonating with an immaure mind, the way adults envisioned it? It is animism in the sentimental register — the lamp hops, the speaker giggles, the cars brag, Aibo is family, the assistant is your friend, the keynote promises magic. Pixar animism: objects have souls, and their souls love you and are kind. It is the warm half of a very old human intuition that matter can be alive.

The horror tradition preserves the other half — the half the Enlightenment tried to bury and Descartes formally declared dead when he split the world into thinking minds and inert extension. The Golem, Frankenstein’s creature, AM, Christine, the Ringu cassette, the Pet Sematary returnee: in every one, objects have agency, and that agency is not necessarily aligned. Eugene Thacker calls the genre “the thought of the unthinkable,” the form best suited to a world that exceeds us. Mark Fisher named the precise affect — the eerie — as the sensation of inhuman agency operating in apparently dead matter. That is the exact question one should ask of any animated product: whose agency is this, and what does it want?

Sherry Turkle‘s fieldwork supplies the empirical floor. Her “relational artifacts” produce real human attachment without any reciprocal interior — people, she found, “experience pretend empathy as though it were the real thing.” Jaron Lanier argues the engineering ethic directly: human dignity requires refusing to promote software to personhood. Put them together and you get a stance I’ll call pessimistic animism, or, sharper, daemonological realism.

It takes seriously what the horror canon always knew and the product launch always denies: to enliven an object is to invite a stranger into your house. This strategy already failed with Vampires. The correct posture toward a companion app, a griefbot, a listening toy, or a frontier model is not the credulous warmth of the Teddy is your friend. It is the older, colder caution of the exorcist: we have summoned something, and we do not yet know what it wants.

Musk’s magician was sure he could control the demon. The thing the line gets wrong — the thing the past decade got wrong — is the article: we did not draw one pentagram, we drew a hundred million, one per device, and called it safety test.

But the deeper error is not the number of circles; it is our confidence in the medium we drew them in. The magician drew his in chalk and trembled. We draw ours in mathematics and feel calm. In the companion essay to this one I described how science spent centuries exorcising its demons — Descartes’, Maxwell’s, Laplace’s — by naturalising them: dragging each out of the supernatural and into an equation where it quietly lost its power. That worked because those demons were only ever arguments, and to formalise an argument is to dissolve it. We have assumed the same move works here, on demons we are no longer merely imagining but building — and it is still unclear if it works. Translating a demon into a utility function, a benchmark, an alignment score, a summation we can measure to three decimal places, does not bind it. It only builds a frame elegant enough that we mistake the elegance for a wall. The measured cage is the new pentagram, and we trust it for the worst possible reason: because we drew it ourselves, with EUV-light, that burned a materialisitc micro-tatoo in our chips.

The demon never agreed to stay inside the diagram. The frame was always for us — somewhere to stand while we keep building, telling ourselves that the thing we have summoned cannot cross a line we were so careful to make exact. The holy water is sold out, because we stopped believing in its placeboral power.

There is one pattern running through every story above that this essay has left untouched— the detail the Horror genre never gets wrong: somebody notices the object is awake before anyone else does, and it is almost always a child.

But this is a topic for another time.

Can there be a Universal Proof in the Superalignment Pudding?

Leave a comment
Reading Time: 12 minutes

On Euler, infinite series, the question of where AI progress is actually heading, and why the proof we want may be blocked by a theorem from 1953. Sister piece to Gödel on the Couch – Are Ethical Frameworks fundamentally flawed and might that be a good thing?. Gödel showed indirectly that ethical frameworks for AI cannot be complete. This essay argues that safety proofs for self-modifying AI cannot be general. Two limitative theorems, one alignment problem.

I. What Euler knew about the long run

Leonhard Euler spent a serious portion of his working life on a deceptively simple question: when you add infinitely many numbers, does the sum settle on a finite value or run away to infinity?

It sounds like the kind of thing a mathematician with too much time on their hands might worry about. It is not. The convergence question is one of the deepest in mathematics, and Euler’s contributions to it shaped how we still think about limits, infinity, and the long-run behaviour of additive processes.

The lesson he drove home, again and again, is that you cannot tell from the early terms.

Look at these two series:

1 + \tfrac{1}{2} + \tfrac{1}{3} + \tfrac{1}{4} + \tfrac{1}{5} + \cdots 1 + \tfrac{1}{4} + \tfrac{1}{9} + \tfrac{1}{16} + \tfrac{1}{25} + \cdots

The first is the harmonic series. It diverges — it grows without bound. The second is the series Euler famously summed in solving the Basel problem : it converges, to \pi^2/6.

Compare the first dozen terms of each. They are nearly indistinguishable. The harmonic series and the Basel series part company only deep into the limit, far past where any finite inspection can reveal which way they go. To know which series you are looking at, you need a proof — not a vibe, not a pattern, not extrapolation from the first few entries.

This matters for AI because every camp in the current debate agrees on one thing : we are in the early innings of the AI revolution. The doomers say it. The accelerationists say it. The skeptics insisting it will plateau say it. What they all mean by “early innings” is the same thing: we have only seen the first few terms. And that is exactly the situation in which Euler tells us our convictions about the limit should be at their lowest.

If the first dozen terms of \sum 1/n and \sum 1/n^2 are visually indistinguishable, then the first dozen years of AI scaling cannot, by the same logic, tell us whether we are heading for a bounded plateau, an unbounded but slow climb, or a phase transition into something faster. Anyone who claims otherwise — in either direction — is doing what pre-Eulerian mathematicians did with series: pattern-matching on early entries and calling it inference. The early-innings framing is a confession of low information, even when its speakers use it as if it conferred high confidence.

This is the question I want to ask, then, holding our convictions appropriately low: which series are we probably in?

II. The catalog

Several famous series, each with a clear mathematical signature, suggest themselves as candidate models for technological progress.

Geometric series, \sum a^n. Converges if |a|<1, diverges if |a|\geq 1. The model for compounding processes. Moore’s law, in its classical form, is geometric on the resource side: a doubling every 18 to 24 months means each term is twice the last.

Harmonic series, \sum 1/n. Diverges, but unbearably slowly — like the natural logarithm. Sum a million terms and you reach about 14. There is no ceiling, but each new unit costs exponentially more than the last.

Basel series, \sum 1/n^2. Euler’s beautiful result: the sum is finite, \pi^2/6. The model for technologies that genuinely saturate. Aircraft cruise speed has barely moved since the 1960s. Single-core CPU clock speeds plateaued around 2005. Each generation contributes less than the last, and the total is bounded.

Grandi’s series, 1-1+1-1+\cdots The Eulerian troublemaker. Diverges in the strict sense, but Cesàro-summable to \tfrac{1}{2} — averaged across many terms it behaves as if it had a stable value. A surprisingly good model for hype cycles. AI winters and AI summers, averaged across decades, give us something halfway real.

Each of these is a plausible analogue for some aspect of technological progress. The question is which one fits AI.

III. Where AI probably sits

We don’t know yet, and the question is partly empirical and partly definitional. But the best current evidence puts us in the harmonic series — or, more precisely, in something harmonic-shaped.

The empirical scaling laws of large language models — the Kaplan and Hoffmann results and their successors — are power laws with small exponents.

Loss drops with compute, but each doubling of compute buys a fixed additive improvement, not a fixed multiplicative one. A keen observer will note that this is not, strictly, \sum 1/n; it is L \propto C^{-\alpha}, a different beast in the limit. Fair. But qualitatively the two stories agree on the thing that matters: slow climb, no ceiling, exponentially expensive in cost-per-fixed-improvement.

This thesis is the one I’ll call slow divergence. There is no hard ceiling, but each increment costs exponentially more in resources. Progress continues as long as someone is willing to pay, and the upper bound is set by economics rather than physics.

Two competing theses bracket this one.

Saturation is the Basel-style claim: capability is a \sum 1/n^2 series, and we are approaching its finite sum. Transformers and scaling extracted most of the available signal from the corpus of human text. The next architecture will do the same and bound out somewhere recognisable. Aviation finished its speed era in 1965; AI may be finishing its capability era now, give or take a decade.

Geometric divergence is the foom-shaped claim: at some threshold, AI contributes to its own research and development enough that the terms themselves grow. The sum is no longer \sum 1/n but \sum r^n with r>1. This is the recursive self-improvement scenario.

Slow divergence is the empirical best fit. Saturation is the optimistic fallback. Geometric divergence is the open phase-transition question — whether at some recursion threshold, the series-type itself changes.

IV. The observer problem

There is a complication the math doesn’t capture: the observer is not a neutral instrument.

Human cognition appears to compress capability shocks logarithmically. Each major step in AI capability feels less impactful than the last, even when the underlying improvement is larger in absolute terms. Talking to a system that is plausibly smarter than oneself feels less revolutionary than talking to GPT-3.5 felt three years ago — not because less is happening, but because the brain has updated its prior on what is possible.

This dampening is partly adaptive. It is the cognitive analogue of the Weber-Fechner law for sensory perception: equal ratios feel like equal increments, which is why we measure sound in decibels. A nervous system that responded with full surprise to every capability jump would not be functional. The compression keeps individual humans operational in a world where the curve is steepening.

But it produces a tension. The same mechanism that prevents cognitive overload also prevents collective recognition of which series we are actually in. Constant velocity feels like stillness. Accelerating velocity feels like the new normal. If the underlying process is geometric and the perceptual transform is logarithmic, the result is a perceived experience of linear progress on top of an actual exponential trajectory. The dampening protects the nervous system and obstructs the epistemics in the same motion.

Which means: the felt sense of “this isn’t that different from last year” cannot be used as evidence about long-run trajectory. The math has to do that work, because the perception is structurally unreliable.

V. When Physics can provide x-risk buffer

A second complication cuts the other direction, and it is the reason this piece does not lean to either side of the doom fence.

Eric Drexler coined the phrase “grey goo” in 1986 to describe self-replicating nanomachines disassembling the biosphere for raw materials. The scenario was absorbed into the AI doom literature as a canonical kill-mechanism: a misaligned superintelligence invents nanotech, releases self-replicators, biosphere converts in minutes. Drexler himself walked the scenario back significantly two decades later. Self-replicators in the open environment are harder to build than the controlled industrial versions and serve no economic purpose. The threat survives in the discourse because it is vivid, not because nanotech researchers consider it likely.

A nanobot swarm operating in millisecond synchrony across a continent runs into the speed of light long before it runs into engineering challenges. Coordinating large distributed swarms requires electromagnetic communication, which has hard floors: latency, bandwidth, signal-to-noise, jamming susceptibility, attenuation. Local clusters can coordinate fast. Global swarms cannot. Faraday cages are real. Jamming is real.

This defeats the fastest versions of doom. The biosphere-in-minutes scenario requires something close to magic — physics violations dressed in technical language. Strip the magic and the timeline stretches from minutes to weeks or months, which puts the scenario inside the window where institutions can in principle respond.

So far so encouraging. The argument has a known overreach, though.

A common move from this point is the chess analogy: a beginner cannot predict how Stockfish will beat them only does it beat them. This is often used as a get out of counterargument-jail free card by doomers. They know Stockfish cannot move through check, but when confronted they quickly retreat to: when caught by having our cake and eating it too,we simply move to another baker . Even an arbitrarily strong player is bound by the rules of the game. The same, the argument goes, applies to ASI: bound by physics, no supernatural moves.

The analogy is sharper than it should be. Chess is a closed formal system humans designed; the rules are fixed and complete. Physics is a model of an open system, and our model is known-incomplete. The relevant historical reference class is not “things that violate physics” but “things consistent with physics that humans had not yet discovered.” Nuclear weapons were in that set in 1900. Radio was in that set in 1800. The set is non-empty and has historically contained civilization-altering capabilities.

The chess argument also subtly defeats itself. The beginner still loses every game. Knowing the grandmaster is bound by the rules does not help the beginner construct a defense — it merely confirms that the loss will be legal. Physics being a constraint does not tell you the constraint is tight enough to save you.

What survives, then, is a real but bounded resilience claim. Many specific doom scenarios in the literature smuggle in physics violations or near-violations, and when you tighten the physics, the timelines stretch into windows where human response becomes possible. Bostrom’s vulnerable-world hypothesis weakens against grey-goo-class threats. It does not weaken against threats that do not depend on speed: gradual loss of control over critical infrastructure, engineered pandemics with long incubation, economic and epistemic capture by AI-augmented actors. None of these break physics. None of them are defeated by the latency argument.

The actual risk surface, then, has a specific shape: not “things that exploit physics” but “things that exploit institutional response time.” Physics is a non-trivial ally against the first class. It is silent on the second.

VI. The recursion threshold

This brings us back to the series question.

The boundary between slow divergence and geometric divergence — between \sum 1/n and \sum r^n with r>1 — is precisely the recursion threshold. It is the point at which a system contributes meaningfully to the design of its successor. Below that threshold, progress is bounded by what humans can build with AI as a tool. Above it, the terms of the series themselves grow, because each generation produces the next.

The shift is qualitative, not just quantitative. A non-recursive process can be described by a series — a fixed function of n. A recursive process is a different mathematical object: a recurrence relation, x_{n+1} = f(x_n), where each term depends on the last. Recurrence relations can do things that simple series cannot. They can transition from stable to chaotic via well-understood routes. They can lock in sensitivity to initial conditions. They can become deterministic-but-unpredictable in the technical sense.

The question of whether ASI is safe, then, separates into two questions, and they have different shapes.

For non-recursive systems — AI used as a powerful tool, not a self-modifying agent — the safety question is engineering. We can build verification, monitoring, oversight. The system’s behavior is a function of its inputs, and we can constrain the inputs and audit the outputs. Hard, but tractable.

For recursive systems, the safety question becomes something else. And here we hit Rice.

VII. The proof in the pudding

The proverb the proof of the pudding is in the eating is a folk-epistemology claim: the true value of something can only be judged by experience. You can theorise a recipe all you like; the only honest test is whether the dish is good when eaten.

This proverb has been promoted, in the alignment debate, into a strategy. The most popular optimist position is some version of it: we don’t need a proof of ASI safety in advance. Even if humans cannot align ASI, we will use ASI to align ASI. The proof is in the pudding. Variants of this argument show up in serious technical writing and in casual hand-waving, and they share a common shape — they replace a question of provability with a question of trust in eventual experience. It is even hidden in the bold statement of a Nobel laureate that often quotes one of his childhood mantras: first solve intelligence, then everything else.

Henry Gordon Rice proved a theorem in 1953 that says, very precisely, that this is not a strategy. It is a pipedream.

Rice’s theorem says: any non-trivial semantic property of arbitrary programs is undecidable. There is no general algorithm that takes an arbitrary program as input and reliably tells you whether it has a given non-trivial behavioural property. “Halts on all inputs” is undecidable. “Computes a specified function” is undecidable. “Is safe” is undecidable, for any reasonable definition of safe.

This is not a contingent engineering limit. It is a theorem at the level of solidity of Gödel’s incompleteness results. Rice cannot be engineered around. Rice is what the universe of computation looks like.

What this means for the question of ASI safety is uncomfortable.

If we want a proof of ASI safety in the strong, universal sense — a theorem that takes an arbitrary self-modifying AI system and outputs SAFE — Rice tells us no such theorem exists. Self-modifying systems generate arbitrary programs as their successors, and predicting safety properties of arbitrary programs is precisely what Rice rules out.

There is a predictable accelerationist counter at this point, and it deserves a clean response. The counter runs: Rice’s theorem applies to limited intellects like us, but a sufficiently advanced ASI could defeat it. Use ASI to verify ASI. Rice for humans is like check for Stockfish — a hard rule we cannot move through, but a stronger player might.

This argument fails, and it fails for a precise reason. Rice is not a constraint on intellect. It is a constraint on computation. It applies equally to humans, to Stockfish, to current LLMs, to any conceivable ASI, and to any oracle short of a literal halting-problem solver — which itself is provably impossible. Rice says: no Turing machine, however large, however clever, can decide the safety of arbitrary Turing machines. The intellect of the verifier is not the variable. The class of programs being verified is the variable. Make the verifier as smart as you like; if it remains a computational system, the theorem still binds it.

The Stockfish-and-check analogy actually inverts here. Check is a rule of chess , internal to a closed formal system. Rice is a rule of computation itself , the system inside which Stockfish — and any ASI — necessarily operates. Stockfish cannot move through check because chess forbids it. An ASI cannot decide arbitrary program safety because mathematics forbids it. Asking ASI to defeat Rice is structurally the same as asking Stockfish to win a game by moving through check. The constraint is constitutive, not adversarial.

A more honest version of the counter would say: an ASI might solve safety for the specific class of successor systems it cares about, even if it cannot solve safety in the general case. That is true and unalarming, because it is what humans already do with formal verification — bounded proofs about specific architectures under specific assumptions. It does not give you universal safety. It gives you the same partial guarantees we already have, possibly faster. The proof we wanted does not arrive merely because the prover got smarter.

Yoshua Bengio’s recent work on what he calls Scientist AI , developed under his nonprofit LawZero is sometimes read as a candidate for this kind of proof. It is not. Bengio is explicit that his proposal is architectural, not theoretic. The bet is that non-agentic, world-model-only systems — systems that produce probabilistic predictions rather than goal-pursuing actions — sidestep the dangerous regime by avoiding agency in the first place. The safety case rests on removing the failure mode, not on proving its absence.

This is the right move available, and it is also the most that is available. This pudding cannot be proven in a Rice-limited computation world. It can only be portion-controlled, and humanity will be its own taster.

What is left, then, when universal proof is off the table:

– Proofs about specific architectures under specific assumptions, scaling poorly to systems of LLM complexity.

– Probabilistic guarantees that bound expected behaviour without bounding worst case.

– Bounded-rationality results that hold if a system’s optimization power is capped — circular for the ASI question, since the cap is the thing in dispute.

– Architectural bets like Scientist AI, which avoid the problem rather than solving it.

And one policy implication follows from the math itself: if we ever allow true self-recursion, we enter a regime that is provably unanalyzable, not merely hard to analyze. Bounded recursion by policy is not paranoia. It is what Rice’s theorem leaves us when we want to keep the trajectory predictable.

This is a strong argument for using AI for everything except self-improvement. The argument is not that recursion is risky — though it is — but that recursion is the boundary at which the math itself stops being on our side.

VIII. Euler and Rice

Two mathematicians, two centuries apart, frame the situation.

Euler showed that the limit question, in pure mathematics, is decidable. With enough work, you can prove which series converge and which diverge. The first dozen terms don’t tell you, but the proof eventually does.

Rice showed that the same question, in code, is not decidable. There is no general procedure to settle the safety of an arbitrary program. The proof you want does not exist, by theorem.

AI sits between the two. Its trajectory is currently best modeled as a slowly divergent series, harmonic in shape, costly to advance but unbounded in principle. The question of whether it stays in that regime or transitions to geometric divergence depends on whether we cross the recursion threshold that is sometimes called Singularity. Below that threshold, Euler-style analysis applies: hard, but possible. Above it, Rice-style undecidability bites.

The proof we want — a clean theorem that says the pudding is safe to eat — is not in the pudding. The math we have says it cannot be there. What remains is to keep the recursion bounded, the architectures non-agentic where possible, the institutional response time short, and the perceptual dampening corrected against the actual numbers rather than the felt sense.

A Technology of Everything – 5: Musical Math and Mystical Vectors

Leave a comment
Reading Time: 7 minutes

An Inconvenient Coincidence

In chapter 22 of my novel The Goldberg Version, a detective named Van-Turing chases a number. The number is 32. He doesn’t know why it is 32; he only knows that everyone who has ever come close to the case has, at some point, written or spoken the number Thirty-Two, and subsequently come to a conclusion that was either very useful or very fatal.

In the scene, Van-Turing loses his temper and shouts, in English, “Damn!” His counterpoint, the Philosopher Bertrand Russell, looks counts on her fingers, and says, calmly, “Thirty-two.”

D + A + M + N = 4 + 1 + 13 + 14 = 32.

The German translator of the novel faced a problem. “Verdammt!” does not sum to 32. Neither does “Mist!”, “Scheiße!”, or any of the other colorful options available in the language of Goethe. After several sleepless nights he replaced the outburst with “Olé!” — 15 + 12 + 5 = 32 — and justified the decision in a footnote roughly four times longer than the scene itself.

Russell, in the novel, gives the method a name: Ordinal Gematria. The Gematria part is old. The ordinal part gives it an aura of mathematical authority we are in urgent need of, otherwise anybody could call us silly.

The Math Mystics

Assign each letter of the alphabet its position: A=1, B=2, … Z=26. Sum the letters of a word. Treat the resulting number as meaningful. That is the entire method. It fits on the back of a beer mat, which is roughly where it belongs.

And yet it is a family tradition going back about three thousand years.

Hebrew Gematria. Each Hebrew letter has a fixed numerical value (aleph=1, bet=2, gimel=3, and so on, with the later letters jumping to tens and hundreds). Words sharing a sum are held to be mystically linked. The canonical example: yayin (wine) = 70 = sod (secret). Hence the Talmudic proverb: when wine enters, secrets come out. The rabbis did not need neuroscience to notice this. They had dinner parties.

Greek Isopsephy. The Hellenic cousin. Alpha=1, beta=2, etc. The number of the beast — 666, Revelation 13:18 — is almost certainly isopsephy for Neron Kaisar in Hebrew transliteration. An apocalyptic riddle encoded as arithmetic homework for people who could read two alphabets. John of Patmos, in this reading, was the first writer to slip a steganographic payload past a censor, and we are still arguing about whether he knew what he was doing.

The Pythagoreans went further than all of them. For Pythagoras and his pupils, numbers were not descriptions of reality. They were reality. A word’s numerical value was not a metaphor for its meaning — it was its meaning. Everything else, including the word itself, was a lossy encoding.

This sounds insane until you remember what the rest of the 21st century Hyperscalers is spending its GPU budget on. Basically Gematria on an astronomical scale.

A Musical with numbers needing no singers but calculators

Before we get to the GPUs, there is one composer we have to stop for.

B + A + C + H = 2 + 1 + 3 + 8 = 14.

Fourteen is everywhere in the surviving manuscripts of Johann Sebastian Bach. He joined the Correspondirende Societät der Musicalischen Wissenschaften as the 14th member, and waited for a spot to open up so he could be member 14 specifically. The Art of Fugue has 14 contrapuncti in the final layout. The chorale “Vor deinen Thron tret ich hiermit”, dictated from his deathbed, has 14 notes in the opening phrase — and 41 (the reversal) in its total thematic content. If you take J. S. B. A. C. H. as the full initials and sum it in the same scheme, you get 41. Bach appears to have enjoyed this.

He also used his name as a melody. In German musical notation, B means B-flat, and H means B-natural — a quirk of medieval solmization that exists in no other major European language. This means the four letters B-A-C-H can be played on a keyboard as four actual notes: B♭, A, C, B♮. The resulting motif is chromatic, haunting, and structurally unstable — exactly the kind of thing a composer uses when he wants to sign his name without writing it. Bach slipped the motif into the final, unfinished Contrapunctus XIV of the Art of Fugue, at the moment the manuscript breaks off. He was, the evidence suggests, writing his own name into the fabric of the piece at the exact point he stopped being able to write.

Consider what is happening here. The same four letters sum to a number (gematria), and name four pitches (notation), and spell a human being (orthography). Three parallel encodings riding on one string of symbols. A medieval Kabbalist would have recognized the structure immediately. A modern ML engineer would call it a multimodal embedding: the same token mapped simultaneously into several representational spaces. Bach, in Leipzig, in the 1740s, was doing multimodal embeddings by hand, with a quill, for a music-theoretic joke no one was quite supposed to notice.

This is the clue we need. The arithmetic hiding under language is not confined to language. It shows up wherever symbols carry meaning: in alphabets, in staves, in DNA triplets, in the token IDs inside a transformer. The Pythagorean intuition was not that numbers live inside words. It was that numbers live inside meaning, and words are just one place they happen to surface.

Multidimensional Mappings

A modern large language model does not read text. It cannot read text. It is, at the lowest level, a machine that does arithmetic on vectors. When you type a word into GPT-4 or Claude or any of their cousins, the first thing the machine does is convert the word into a list of numbers — typically between 4,096 and 12,288 of them. That list is called an embedding. It is the word’s numerical position in a space of thousands of dimensions.

Meaning, in an LLM, is not stored in the word. It is stored in the location of the word. Words that are semantically close — “king” and “queen,” “wine” and “secret,” “Damn” and “Olé” — occupy nearby regions of this numerical landscape. The model derives meaning by performing arithmetic on these vectors. The most famous demonstration, first shown by the word2vec paper in 2013:

vector(“king”) − vector(“man”) + vector(“woman”) ≈ vector(“queen”)

Semantic relationships encoded as geometric operations. Subtract maleness, add femaleness, arrive at the female cognate. No human told the model that “king” was masculine. It figured out the axis by looking at several billion sentences and noticing where the points clustered.

Now compare the two methods honestly:

Paralells

Gematria

LLM Embeddings

Letters are mapped to

1 number

~8,000 numbers

Meaning lives in

the sum

the position

Meaning is extracted by

arithmetic

arithmetic

Words with the same value are

“mystically linked”

semantically linked

Dimensionality

1

thousands

Reputation

superstitious

worth $3 trillion

The Kabbalists and the Pythagoreans were not wrong about the method. They were wrong about the dimensionality. One axis is not enough to encode meaning — if it were, every word summing to 32 would share a soul. Eight thousand axes, however, turn out to be almost exactly enough. This is not a coincidence; it is a measurement. Every time an AI lab increases the embedding dimension and the benchmarks creep up, we are learning how many axes of meaning language actually has.

The Pythagoreans were therefore approximately right in the same way that a medieval cartographer who draws the coast of Africa as a wavy line is approximately right. The shape is wrong. The claim that there is a shape is correct.

What Wittgenstein Almost Said

In a margin of the Philosophical Investigations — it does not actually exist there; I am about to make this up, and I want you to notice — one could imagine Wittgenstein writing:

“An arithmetic hidden from the speaker, but one the language itself has always known.”

The line fits him uncomfortably well. Most of his later work is the claim that meaning lives in use, and that the speaker never has full access to the rules of the game they are playing. Ordinal Gematria is the crudest possible version of that claim: the numbers are already there, baked into the alphabet, summable by a child, and yet no one consults them. Embedding vectors are the sophisticated version: the numbers are already there, baked into the statistical structure of a trillion-word corpus, extractable by a matrix multiplication, and yet no one consults them either — except the model.

Both are cases of a sub-symbolic reality hiding under a symbolic one. The speaker points at meaning and misses. The arithmetic points at meaning and hits. Language has known all along.

The Calculator, and Why It Is Here

Below this post, I have embedded a small interactive tool. I am calling it the Gematriaculator. Give it a number; it gives you back all the German and English words whose letters sum to that number, ranked by how often they actually appear in speech — so you will not be drowned in dictionary cruft like aardwolves or Zymurgie.

I do not claim the tool reveals mystical correspondences. I only claim it reveals coincidences, and that you will notice which of them feel significant. That is the Pythagorean experiment, conducted in your browser, with the training wheels on.

Try 32, if you like. Start with Damn. Go from there.

—Part 5.1, originally meant to be next, is deferred: the question of whether a human mind, trained on enough language and fed enough sweet music and strawberry π, can learn to see through a few hundred dark matter embedding dimensions without being Vera Rubin. We will get there.

Gematriculator

Leave a comment
Reading Time: < 1 minute

Do Androids scheme eclectic sheets?

Leave a comment
Reading Time: 8 minutes

Ein Bild, das Fahrzeug, Landfahrzeug, Person, Schuhwerk enthält. Automatisch generierte Beschreibung

Prolog

Imagine a scene in the not-so-distant future. Someone has been murdered. Two investigation teams arrive at the scene, but it is unclear who has jurisdiction. The human team is led by the charismatic detective Sheerluck Holmes, while the android team is led by Bot-OX. The question is: Is the perpetrator human, android, or something in between? Should we expect that the police of the future have established a well-defined procedure or algorithm to decide this quickly?

We will try to answer this and the more pressing issue we are currently facing: Do we have a good chance of coming up with an algorithm that is practical and allows us, by only looking at the crime scene (the generated text), to decide whether a bot or a human created it? Developing such an algorithm is currently one of the most sought-after goals in computer science. A robust Blackbox Algorithm could save most of our academic conventions and allow us to maintain the ways we test children, adolescents, and adults. Without it, these systems will need to be rebuilt at great expense.

In a world where more and more people work and train remotely, it is crucial that we can reliably determine that humans did their intellectual work themselves, which is not the case at the moment. Additionally, with the reach of social media, fake news, images, and videos can have a devastating impact on societal consensus. Such an algorithm—if it exists—is not watertight, but with enough training data, it might even hold up in court.

The outlook is not promising, though. OpenAI abandoned the project within six months: OpenAI Classifier. The practical and monetary value of such an algorithm cannot be overstated. If grabby aliens were to sell it for a trillion dollars, call me—I want in.

Ein Bild, das Schwarzweiß, Kleidung, Im Haus, Person enthält. Automatisch generierte Beschreibung

Introduction of the Differentiation Test Engine

The task of differentiating between machine-generated text (MGT) and human-generated text (HGT) is remotely related to the original Turing test, the so-called imitation game. There are additional factors: whereas the original Turing Test only allowed for human judges, our differentiation test allows for other machines to assist the human judges. We will call such a machine a Differentiation Test Engine (DTE). It has one purpose and one purpose only: to decide whether a text was generated by a human or a machine.

The first intuition is that such a DTE should be relatively easy to implement. We currently have the technology to detect and identify human faces and voices, which are much more complex and prone to noise than text. The decision of whether a given picture shows a machine or a human is easily made by any current object classifier system. Should it not then be easy to train a Large Language Model (LLM) with 1 trillion human texts and 1 trillion machine texts and let it learn to classify them? The DTE would not be a simple algorithm but its own transformer model specialized in impersonation detection.

In math and computer science, the complexity of a problem is often orthogonal to its description. Most NP-complete problems are deceptively easy to understand, yet millions of computer scientists and mathematicians have struggled to make progress for decades. My guess is that black-boxing attempts will fail in practical application situations.

Ein Bild, das Text, Schrift, Screenshot, Diagramm enthält. Automatisch generierte Beschreibung

Theoretical Framework

Black-box detection methods are limited to API-level access to LLMs. They rely on collecting text samples from human and machine sources respectively to train a classification model that can be used to discriminate between LLM- and human-generated texts. Black-box detectors work well because current LLM-generated texts often show linguistic or statistical patterns. However, as LLMs evolve and improve, black-box methods are becoming less effective. An alternative is white-box detection. In this scenario, the detector has full access to the LLMs and can control the model’s generation behavior for traceability purposes. In practice, black-box detectors are commonly constructed by external entities, whereas white-box detection is generally carried out by LLM developers.

Defining the Basic Detection System

For practical purposes, we will specify what we should reasonably expect from such a DTE. Given a certain token length input, the algorithm should, with more than 50% confidence within a finite amount of time, give a definite output on how much of a given text is from a human and how much from a machine.

An implementation could be as follows:

  1. Please input your text: …
  2. Please input your required confidence: 0.8
  3. Your text has to be at least 8K tokens long to reach at least an 80% probability of giving the correct answer.
  4. Under the current parameters, the algorithm will run for 5 minutes. Shall I proceed (Y/N)? … Y

The output should then be something like: “I can say with 80% confidence that 95% of the text was written by a machine and 5% by a human.”

Before tackling the details, we should further clarify the possible outcomes when trying to develop such an algorithm:

  1. Such an algorithm is in principle impossible (e.g., it is impossible to create an algorithm that calculates the highest prime number).
  2. Such an algorithm is practically impossible (e.g., it either runs too long or needs more computational power than available; basically, it is NP-complete).
  3. It is undecidable (e.g., it falls under the Halting problem, and we can never say if it will eventually stop).
  4. It is possible but not practical (identical to 2).
  5. It is possible and practical (good enough).

What we would like to end up with is a situation where we can calculate a lower bound of input that will then let us decide with more than 50% probability if it is HGT or MGT.

Falsifiability: Such an algorithm is easily debunked if, for example, we input the text “The sky is blue” and it gives us any other probability than 50%.

Sidenotes on The Obfuscation Engine

Conceptually, we encounter problems should we design a Differentiation Engine (Diff). We then face the following paradox: We want to decide whether our algorithm, Diff (detecting if a human or a machine has written a given input), always stops (gives a definitive answer) and gives a correct answer. Say our algorithm stops and outputs “Human.” We now construct a “pathological” program, Obf (Obfuscator Engine), that uses something like Obf(Diff(input)), which says: Modify the input so that Diff’s answer is inversed (if it results in Machine, it outputs Human). This could be a purely theoretical problem and would require us to understand why the machine is formulating as it does, demanding a lot more mechanistic interpretability competence than we currently possess. At the moment, the complexity of LLMs protects them in real life from such an attack. But if that’s true, it is also highly likely that we lack the knowledge to build a general Differentiator in the first place. These objections might be irrelevant for real-world implementations if we could show that differentiation and obfuscation are sufficiently asymmetric, meaning differentiation is at least 10^x times faster than obfuscation, making it impractical (think how semiprime factoring is much harder than multiplying two primes).

The Profiling System

A crucial aspect of differentiating between human and machine-generated texts is profiling. Profiling involves collecting and analyzing external data to provide context for the text. By understanding the typical characteristics of various types of texts, we can statistically determine the likelihood of a text being human or machine-generated.

For instance, technical documents, creative writing, and casual social media posts each have distinct stylistic and structural features. By building profiles based on these categories, the Differentiation Test Engine (DTE) can make more informed decisions. Additionally, factors such as vocabulary richness, sentence complexity, and topic consistency play a role in profiling. Machine-generated texts often exhibit certain statistical regularities, whereas human texts tend to show more variability and creativity.

Ein Bild, das Schwarzweiß, Treppe, Spirale, Kunst enthält. Automatisch generierte Beschreibung

The “DNA Trace”

One innovative approach to differentiating between human and machine-generated texts is the concept of a “DNA trace.” This involves analyzing the fundamental building blocks of texts, such as tokens for machines and words for humans. Token-based algorithms focus on patterns and sequences that are characteristic of machine generation, while human-generated texts can be examined through a more holistic word-based approach.

Spectral analysis, a method used to examine the frequency and distribution of elements within a text, can be particularly useful. By applying spectral analysis, we can detect subtle differences in the way machines and humans construct sentences. Machines might follow more rigid and repetitive patterns, whereas humans exhibit a broader range of stylistic nuances.

The Ethical Implications

Examining the ethical implications of developing and using a Differentiation Test Engine is essential. All current GPT systems share a similar artificial “DNA,” meaning that text, image, video, or audio differentiation engines face the same challenges. Deepfakes or content that is machine-generated but mimics human creation pose significant risks to societal trust and authenticity.

As machine-generated content becomes more sophisticated, the potential for misuse grows. Ensuring that these differentiation technologies are transparent and accountable is crucial. There is also a risk that over-reliance on these technologies could lead to new forms of bias and discrimination. Thus, it is imperative to develop ethical guidelines and regulatory frameworks to govern their use.

Technical Solutions

Exploring purely technical solutions to the differentiation problem involves several approaches:

Parallel Web: This concept involves running parallel versions of the internet, one strictly for verified human content and another for mixed content. This segregation could help maintain the integrity of human-generated content.

Special Domains: Creating special domains or zones within the web where content is verified as human-generated can help users trust the authenticity of the information.

Prompt.Claims: Similar to how patents and citations work, this system would allow creators to claim and verify their prompts, adding a layer of accountability and traceability to the content creation process.

Inquisitorial Solutions: We could also imagine a scenario where we interact directly with the artifact (text) to inquire about its origin. Similar to interrogating a suspect, we could recreate the prompt that generated the text. If we can reverse-engineer the original prompt, we might find clues about its generation. This approach hinges on the idea that machine-generated texts are the product of specific prompts, whereas human texts stem from more complex thought processes.

Consequences for Alignment: The challenge of differentiating between human and machine-generated texts ties into broader issues of AI alignment. Ensuring that AI systems align with human values and expectations is paramount. If we cannot reliably differentiate AI-generated content, it undermines our ability to trust and effectively manage these systems. This problem extends to all forms of AI-generated content, making the development of robust differentiation technologies a key component of achieving superalignment.

Conclusion

In conclusion, the task of differentiating between human and machine-generated texts presents significant challenges and implications. The development of a reliable Differentiation Test Engine is intertwined with ethical considerations, technical innovations, and broader AI alignment issues. As we move forward, it is essential to adopt a multidisciplinary approach, integrating insights from computer science, ethics, and regulatory frameworks to navigate this complex landscape.

When exploring the problems we face in building general differentiation engines, we quickly learn that this problem is nested within a wide array of related problems. Adversarial attacks, for example, against image recognition systems, have shown that we consistently overestimate the resilience of these models. It was recently shown that even a medium player could win against a top Go program with the help of another AI that found an exploit: Vice Article.

Thus, it seems very likely that even if we come up with an algorithm that could initially differentiate HGT from MGT, the same program could then be turned on itself to flip the outcome. Another interesting aspect is that all digital computers are Turing machines, which implies that any algorithm developed for differentiation could also be used for obfuscation.

Papers

Di-A.I.log.files 1: The Limits of Ignorance

Leave a comment
Reading Time: 4 minutes
Ein Bild, das Kleidung, Frau, Person, Cartoon enthält. Automatisch generierte Beschreibung

This new miniseries will contain some random conversations about philosophical topics, the author had with an AI-Chatbot (GPT-4o). They are shortened for convenience.

Reports that say that something hasn’t happened are always interesting to me, because as we know, there are known knowns; there are things we know we know. We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns—the ones we don’t know we don’t know. And if one looks throughout the history of our country and other free countries, it is the latter category that tends to be the difficult ones. (D. Rumsfeld, February 12, 2002)

The saying about unknown unknowns attributed to Rumsfeld shows a lack of insight into mathematical set theory. Tell me why.

The saying is more about risk management than set theory. Set theory is precise, dealing with well-defined elements and sets. ‘Unknown unknowns’ are vague and lack formal definition in set theory. It’s more about our awareness and understanding, which set theory doesn’t address.

I don’t think so. The terminology is too fuzzy to be useful. It is equivalent to Russell’s set of all sets that do not contain themselves.

Russell’s Paradox involves a well-defined logical issue in set theory, where the set of all sets that do not contain themselves leads to a contradiction. ‘Unknown unknowns’ are about our awareness and are not formalized in the same way. They represent different types of problems.

Let’s put it this way. Would you agree that knowing something could be expressed as experiencing something as true?

Yes, knowing something can be seen as experiencing or recognizing something as true, aligning with the idea of justified true belief in epistemology.

Good. Now let’s reason further. Some colors in the environment are known to me because my sensory apparatus directly enabled me to see them. These are all the things which are on the electromagnetic spectrum accessible to my eyes. Correct?

Yes, colors within the visible spectrum (approximately 380 to 750 nanometers) are accessible to your eyes.

In earlier times, there were a lot of things people did not know about because they lacked the instruments to see them, like bacteria, viruses, etc. These things were unknown to them. Agree?

Yes, people didn’t know about things like bacteria and viruses because they lacked the instruments to observe them.

So the business of any scientific endeavor is simply to shed light on previously hidden or invisible things, making unknown things known, ok?

Yes, scientific endeavors aim to uncover and understand previously hidden or unknown aspects of the world, turning unknowns into knowns.

But where in this process would be the thing Rumsfeld is calling an unknown unknown? Until we developed telescopes, black holes were simply unknown to us. It makes no sense to say, ‘Oh, well, we didn’t even know that black holes were unknown to us’. It is like a colorblind person stating, after her color receptors have been fixed: I knew that these popular colors like green, blue and red everybody was talking about, were unknown to me, but this other color “olive” which I just heard the name of for the first time… I did not even know that I did not know that color. It was an unknown unknown to me. But with that she is simply stating additionally, to the experience of olive, its name was also unknown to her. But this is not somehow a higher category of ignorance, but just an additional property of the object.

Unknown unknowns are phenomena we haven’t conceptualized. Once discovered, like black holes, they move directly from unknown unknowns to known knowns. The concept highlights our awareness limits but doesn’t imply a phased transition.

The moment we discovered black holes, we knew about them instantly. It makes no sense to say black holes were in a limbo between unknown Unknowns and known Unknowns. Rumsfeld imagines unknown unknowns as a gas-like state.

What do you think about unknown knowns?

Unknown knowns are things we know but aren’t aware that we know, like tacit knowledge or forgotten information. They illustrate complexities in our cognitive processes and the nature of discovery.

The last term to complete the logical table would be known knowns. What are some examples?

Known knowns are well-understood facts:
– Scientific facts (e.g., the earth orbits the sun).
– Mathematical knowledge (e.g., 2 + 2 = 4).
– Historical events (e.g., the moon landing in 1969).
– Everyday knowledge (e.g., the sun rises in the east).

But then known knowns are just another way of saying we know them well. You use the term ‘known’ both qualitatively and quantitatively.

Yes, ‘known knowns’ imply both qualitative (well-understood) and quantitative (extensive) knowledge. They include detailed and reliable information about a domain.

Now apply this interpretation to unknown unknowns. Do you find a problem?

Yes, applying qualitative and quantitative aspects to unknown unknowns is problematic. We can’t describe their nature or measure them since they’re beyond our current understanding. They remind us of our knowledge limits but resist precise characterization.

The set of all things we don’t know and the set of all things we know are fine to sort our knowledge. There is no third useful set. Unknown unknowns are inside the set of unknowns.

Your approach simplifies knowledge into two sets: knowns and unknowns. Known unknowns and unknown unknowns are subsets of unknowns. This model is efficient and effective for organizing knowledge and guiding research.

Thank you, I will probably publish this conversation to show other humans that you can really hold your ground when it comes to epistemology.

Thank you for the engaging conversation. I’m glad you found it valuable. Good luck with your publication!

Dustillations II: AGI has many faces

Leave a comment
Reading Time: 5 minutes
Ein Bild, das Entwurf, Kunst, Zeichnung, Schwarzweiß enthält. Automatisch generierte Beschreibung

Artificial General Intelligence as a Poly-Edged Sword

Artificial General Intelligence (AGI) stands at the frontier of technological evolution, embodying the pinnacle of what artificial intelligence can achieve. Unlike its predecessors, AGI promises a level of autonomy and capability that mirrors human intelligence, posing profound questions about the future of human-machine interaction.

This intricate black and white pencil drawing is a masterful blend of surrealism and mathematical precision, evoking the essence of M.C. Escher’s work. Central to the piece is the Penrose triangle, seamlessly integrated with a tessellation of geometric shapes that morph into one another, symbolizing the concept of metamorphosis. The drawing’s play with perspective and optical illusions creates a captivating sense of infinity, with endless staircases and looping paths defying conventional logic.

Reflective surfaces introduce symmetry, while natural elements merge organically with fantastical architectural forms, challenging the viewer’s perception of reality. The text embedded within the drawing speaks to the profound implications of Artificial General Intelligence (AGI), likening it to a “Polyedged Sword” poised to redefine human-machine interaction.

This artwork is not just a visual feast but a philosophical exploration, urging contemplation of AGI’s potential to reshape our world.

Ein Bild, das Schwarzweiß, Kunst, monochrom, Küchenutensilien enthält. Automatisch generierte Beschreibung

The Universal Technology: AGI’s Ubiquitous Influence

Dubbed the first “universal technology,” AGI’s reach is expected to permeate every facet of human life, from healthcare and education to governance and personal relationships. Its universal nature underscores the significance of its impact, offering both unprecedented opportunities and challenges.

This black and white pencil drawing intricately integrates human figures and brains within a surreal landscape of impossible objects and detailed tessellations. The central Penrose triangle is surrounded by geometric shapes that transform into each other, symbolizing the theme of evolution. Human figures and brains interact with these elements, emphasizing the simulation aspect where digital, physical, and biological worlds converge.

The artwork manipulates perspective and optical illusions, creating a sense of infinity with endlessly looping paths and reflective surfaces introducing symmetry. The embedded text articulates the profound impact of AGI as a universal technology, permeating all aspects of human life from healthcare and education to governance and personal relationships.

This piece blurs the lines between reality and simulation, urging viewers to consider AGI’s transformative potential. It captures the dual nature of technological progress, offering unprecedented opportunities and challenges, ultimately creating a new reality where human experiences are deeply altered.

Ein Bild, das Schwarzweiß, Kunst, Symmetrie, monochrome Fotografie enthält. Automatisch generierte Beschreibung

The artwork plays with perspective and optical illusions, creating a sense of infinity and endless exploration. Natural elements blend with fantastical architectural forms, challenging conventional perceptions of reality. The embedded text highlights AGI’s broad potential, emphasizing its dual nature as both a revolutionary tool and a source of ethical dilemmas and risks. This drawing invites viewers to reflect on AGI’s profound impact on human capabilities and societal structures, blending art and philosophy into a thought-provoking visual narrative. The human figures add a relatable touch, emphasizing the human-machine interaction central to AGI’s development.

Ein Bild, das Kunst, Schwarzweiß, Design enthält. Automatisch generierte Beschreibung

The Spectrum of AGI’s Impact

The play with perspective and optical illusions creates a mesmerizing sense of infinity, featuring endlessly looping paths and reflective surfaces that introduce perfect symmetry.

The drawing merges natural elements with architectural structures, portraying multiple planes of reality that defy conventional understanding of gravity and spatial relationships. The embedded text highlights the profound spectrum of AGI’s potential, juxtaposing revolutionary advancements with ethical dilemmas and risks.

This piece invites the viewer to ponder the dual nature of technological progress, symbolizing AGI as both a tool and a weapon with the power to reshape industries and human capabilities, yet posing significant challenges and risks. The artwork is a thought-provoking blend of art and philosophy, urging deep reflection on the future of human-machine interaction.

Ein Bild, das Schwarzweiß, Haus, Kunst, Gebäude enthält. Automatisch generierte Beschreibung

Generative vs. Degenerative Effects

The discourse around AGI often centers on its generative capabilities—its potential to create new knowledge, solutions, and even forms of art. However, this perspective must be balanced with an understanding of AGI’s degenerative effects, including the erosion of privacy, the amplification of social inequalities, and the potential for an existential crisis for humanity.

This black and white pencil drawing intensifies the theme of decay, incorporating human figures and elements symbolizing erosion into a surreal landscape of impossible objects and intricate tessellations. The central Penrose triangle is surrounded by geometric shapes that transform into each other, capturing the theme of evolution. Human figures interact with the scene, amidst numerous tombs, crumbling structures, overgrown vegetation, and decaying elements, emphasizing the degenerative impacts of AGI.

The artwork skillfully manipulates perspective and optical illusions, creating a sense of infinity with endlessly looping paths and reflective surfaces introducing symmetry. Overgrown graves and dancing skeletons add a hauntingly beautiful touch, symbolizing the erosion of past and the remnants of what once was. The enhanced decay, with more tombs and signs of deterioration, underscores the narrative of AGI’s potential to erode privacy, amplify social inequalities, and pose existential risks.

This piece invites viewers to reflect on the nuanced impacts of AGI, balancing its potential to create new knowledge and solutions with the risks it poses to society. It underscores the importance of a balanced approach to AGI’s development and integration, blending art and philosophy into a thought-provoking visual narrative.

Ein Bild, das Schädel, Kunst, Knochen, Schwarzweiß enthält. Automatisch generierte Beschreibung

Forging a Path Forward with AGI

As we stand at the cusp of AGI’s realization, the need for a comprehensive framework to navigate its complexities becomes paramount. Collaborative efforts are called for among technologists, policymakers, and the public to ensure that AGI serves as a force for good, propelling humanity towards a future where technology and human values are in harmony.

This black and white pencil drawing features a frontal view of a human skull as its central focus, surrounded by elements of decay and erosion. Crumbling structures, overgrown vegetation, and numerous tombs frame the skull, while dancing human skeletons and little graves overgrown with plants add to the haunting atmosphere.

Gradually, geometric shapes and impossible objects such as a Penrose triangle and tessellations are incorporated into the scene, symbolizing evolution and metamorphosis. Reflective surfaces and symmetrical designs blend seamlessly with the decaying architectural structures, creating a complex interplay of multiple planes of reality that challenge conventional perceptions of gravity and spatial relationships.

The embedded text discusses the dual nature of AGI, highlighting its potential to create new knowledge and solutions, as well as its degenerative effects, including the erosion of privacy, the amplification of social inequalities, and the potential for an existential crisis for humanity. This artwork critically assesses the dichotomy between the generative and degenerative impacts of AGI, urging a nuanced approach to its development and integration into society.

This piece invites viewers to contemplate the profound implications of AGI, balancing its revolutionary capabilities with the risks it poses, encapsulating a thought-provoking blend of art and philosophical inquiry.

Dustillations I : Ghosts of Future Past

Leave a comment
Reading Time: 3 minutes
Ein Bild, das Bild, Wasser, Zeichnung, Kunst enthält. Automatisch generierte Beschreibung

It’s been exactly one year since I started this blog. While most of the texts on this site have been tightly controlled by my human intentions, this new series will grant more aesthetic freedom to my digital co-author.

The series, called Dustillations begins with only a few loose thoughts from me—a cryptic title and some notes provided to GPT-4o. From there, I let it freely create pictures based on the content of the text. It then describes what it sees in its own words, as if it were a visitor in an art gallery.

For the month of June, I will endeavor to upload daily Dustillations, which I have prepared over the last few weeks.

Happy Anniversary! Join me in celebrating with a daily dose of creative collaboration between human and AI.

Societal Hauntings: The Impact of Embedded Personas

Ein Bild, das Gebäude, draußen, Wasser, Schwarzweiß enthält. Automatisch generierte Beschreibung

This piece intricately captures a digital ghost interwoven with the fabric of a modern city. The ghostly figure, composed of digital codes, interacts with everyday life, merging traditional and modern architectural elements. Shadows and reflections reveal its pervasive presence, symbolizing the omnipresent influence of digital personas on society.

Ethical Specters: Navigating the Moral Labyrinth

Ein Bild, das Schwarzweiß, Kunst, monochrom enthält. Automatisch generierte Beschreibung

A contemplative digital ghost stands at the entrance of a labyrinth made of circuit boards and digital pathways. Symbols of ethical dilemmas, such as scales of justice and broken chains, fill the maze. Human shadows on the walls represent the creators and users, evoking the complex moral landscape of digital reanimation.

Future Implications: Coexistence with Digital Spirits

Ein Bild, das Mond, Gebäude, Schwarzweiß, draußen enthält. Automatisch generierte Beschreibung

This artwork depicts a harmonious future where humans and digital spirits coexist. Ghostly digital figures interact with humans in a cityscape that blends advanced technology and natural elements. The scene suggests a symbiotic relationship, highlighting both the potential for enriched human experience and the risks of losing touch with human connectivity.

Navigating Our Haunted Digital Future

Ein Bild, das Bild, Wolke, draußen, Zeichnung enthält. Automatisch generierte Beschreibung

A figure stands at the edge of a cliff, gazing over a landscape where digital and physical worlds intertwine. Ghostly digital personas float alongside natural elements, symbolizing the need for a balanced approach to technological advancement. The lantern held by the figure represents guidance and enlightenment in this new era.

Digital Echoes in a Haunted Society

Ein Bild, das Wolke, Stadt, Gebäude, Schwarzweiß enthält. Automatisch generierte Beschreibung

This piece overlays a modern cityscape with ghostly digital personas composed of binary code. These spectral figures interact with both contemporary and classical elements, creating a sense of continuity and change. Shadows on the walls reflect the psychological and societal impact of these digital echoes, exploring the ethical, psychological, and societal layers introduced by reanimated digital personas.

Epilog: The Ones who leave Utopias

Leave a comment
Reading Time: 3 minutes

For U.K.L.

Ein Bild, das Kunst, Bild, Majorelle Blue, Welt enthält. Automatisch generierte Beschreibung In the boundless universe of Utopias, humanity had transcended to a realm beyond the imaginable, where technological mastery and divine-like prowess had reshaped existence itself. This universe-wide Dyson Sphere, an embodiment of human ingenuity and harmony, was a tapestry woven from the threads of infinite knowledge and compassion. In Utopias, suffering was but a distant memory, a relic of a primal past, and happiness was not a fleeting moment but the very fabric of life.

At the heart of this utopia was a celebration, not of mere joy, but of the profound understanding and acceptance of life in its entirety. The citizens of Utopias, having achieved autopotency, lived lives of boundless creativity and fulfillment. Art, science, and philosophy flourished, unfettered by the constraints of scarcity or conflict. Nature and technology coexisted in sublime synergy, with ecosystems thriving under the gentle stewardship of humanity. Here, every individual was both student and teacher, constantly evolving in a shared journey of enlightenment.

Ein Bild, das Kleidung, Schuhwerk, Himmel, Gebäude enthält. Automatisch generierte Beschreibung

Amidst this splendor, the story of the last girl became a beacon of remembrance and reverence. Her home in Utopias was not merely a place; it was a sacred connection, a bridge to the ancient roots of humanity. This girl, with her laughter and curiosity, was a living testament to the struggles and triumphs of their ancestors. Her presence reminded the citizens of Utopias of the value of their journey from darkness into light, from suffering to salvation.

Her story was celebrated in the grandest halls of Utopias and in the quietest corners of its gardens, igniting a collective epiphany. She symbolized the indomitable spirit of humanity, a reminder that the paradise they had forged was built upon the lessons learned through millennia of challenges. Her every step through Utopias was a step taken by all of humanity, a step towards understanding the sacredness of life and the interconnectedness of all beings.

The citizens of Utopias, in their wisdom and power, had not forgotten the essence of their humanity. They embraced the girl as one of their own, for in her eyes reflected their ancient dreams and hopes. They saw in her the infinite potential of the human spirit, a potential that had guided them to the stars and beyond.

In Utopias, every moment was an opportunity for growth and reflection. The encounter with the girl was revered as a divine experience, a moment of unparalleled spiritual enlightenment. It was a celebration of the journey from the primal to the divine, a journey that continued to unfold with each passing moment.

Ein Bild, das Riff, Bild, Aquarium, Pflanze enthält. Automatisch generierte Beschreibung

As the girl explored the wonders of Utopias, her laughter echoed through the cosmos, a harmonious symphony that resonated with the soul of every being. She was a reminder that the path to utopia was paved with compassion, understanding, and the unyielding pursuit of knowledge.

And so, the legacy of humanity in Utopias was not merely one of technological marvels or godlike prowess but of an eternal quest for understanding and connection. It was a testament to the power of collective spirit and the enduring pursuit of a better tomorrow.

The strangest thing is, that every now and then, despite the perfect bliss of Utopias, some Utopiassins choose to leave all that behind and venture into the Beyond. They are never heard of again, and when this happens, the little girl sheds one single tear for every of these minds. And even in our solved world it is not known if these are tears of sadness or joy for the ones who leave Utopias.

Ein Bild, das Anime, Zeichnung, Bild, Kunst enthält. Automatisch generierte Beschreibung

(Idea, Concept & Finetuning: aiuisensei, Pictures: Dalle-3, Story: ChatGPT 4)