A new religion emerges
On controlling what cannot be rationally controlled
Trendy as they are right now, this isn’t going to be one of those essays where at the end I tell you I’ve become a Christian. But I’m still going to start it the way most of them do—by telling you just how much of an atheist I once was.
There were no “rules” per-say in my household growing up about how to talk about religion, but if you didn’t want to get laughed at, it was best to stick to talking about it like an anthropologist. Religion was something you observed in other people and other cultures as an objective, disinterested third-party. Obviously, religion was a thing that existed in the world and therefore affected us and required discussing on occasion, but always with the implicit understanding that religion was for “other” people—particularly those that we did not know and did not associate with—a curio to be examined in those poor, primitive souls lacking the wisdom required to wake up to the fact that it’s all a bunch of crap.
I spent the years from Kindergarten to 5th grade in Catholic school, which my mother made sure to clarify was only because the class sizes were smaller than those at the public school, and because the school itself was walking distance from our house. It was not, she insisted it be known, related to the fact that she was raised Catholic, nor to the fact that prior to her disavowal of the faith, she was regularly attending mass at same church I was going to school at.
No, we were atheists. Diehard atheists. The kind that read books about atheism and then try to get everyone else to read those books too. You might have called us evangelical atheists. And as such, atheism became the lens through which I saw the world, and for most of my life, it seemed insane not to be an atheist. How, in this day and age where information is so readily available, where the real world is so easy to comprehend in scientific and mathematical terms, and where the arc of history was so clearly pointing toward irreligiousness, could anyone still willingly ignore all that in favor of letting their entire lives be governed by arbitrary folk tales? How was it possible that people who otherwise behave rationally in nearly all aspects of their lives were capable of entirely suspending that rationality wherever it conflicted with their fairytales?
Though most people probably didn’t grow up with quite the level of militant atheism I did, my experience was hardly unique. Most of my peers had families in either the “recently lapsed” category or the “still absentmindedly performing empty rituals” category, which meant we were all floating in the same direction: toward a version of life governed exclusively by rationality, free from the arbitrary constraints of religious doctrine. To pursue science and to behave rationally, we understood, was to move forward, and anything else, the opposite. The story of Galileo’s persecution by the Catholic Church for proposing what he had rationally deduced—that the earth revolved around the sun—was told and retold throughout my youth and education, serving as definitive proof that religion could be nothing but the enemy of truth.
Such was our indoctrination, and then we were sent out into the world, freed of the superfluous and unnecessary traits that once held us back, and asked, “What great new peaks of human achievement and progress will we now summit?” Surely, this could only lead to some kind of utopia?
The concept of the sacred vanished. Nothing was exempt from the chopping block. “Disrupting” societal norms became a good thing. The economy became our god, and the saints were its agents of capital efficiency—people who “moved fast and broke things.” To participate in “culture”, it turned out, was just a deviation from the path of perfect efficiency. Form became subservient to function. Art merged with politics. We replaced cathedrals first with rec centers, then with Zoom rooms. Communities became apps, and our obligations as members became monthly subscription fees. Weddings were shortened to fifteen minutes as we cut all the boring ritualistic parts out. Satellites flooded the night sky, commingling with the stars. Leather shoes became rubber, then plastic, rebranded for consumers as “vegan”. Wifi appeared on airplanes. Wisdom and folk remedies became WebMD and an army of “research-backed” pills and substances. Food became calories and grams and macronutrient profiles. Eggs became a commodity produced either by sick birds in a warehouse or at a mung-bean processing facility, depending on which definitely-not-a-religion you subscribe to. And then came AI, and finally, the system we’d worked so hard to build had found a way to manufacture art and culture in the very same factories that now produced everything else. We called it “democratization“ and we declared that it was good.
No longer would the intellectual domain be dominated by the few souls lucky enough to posses the talent, and the wisdom, and the perseverance to create great and meaningful things. No longer would art and knowledge work be immune to subjugation by the factory system. No, with AI, even those problems that were complex and unable to be reduced to a single discrete solution came under our control. They could be systematized and incorporated into the machine. Finally the intellectual fruits of humanity, those great and meaningful things, could be manufactured as commodities, on a dollars in, content-out basis, efficiently and rationally.
—
From the dawn of civilization through to today, the horizon of human society, our understanding of things beyond just ourselves and the people we physically encounter, has done nothing but expand. In the beginning we were small tribes of hunter-gatherers, whose sense of the world did not extend far beyond the tribe. But as writing systems developed, a new type of knowledge emerged--one that transcended mere personal experience. Now we could tell stories that could not only withstand the weathering of time, but which could also travel, and be exchanged, such that we could know not just the stories we told of ourselves, but also the stories others told. And the more these stories got written, and the more of them we encountered, the richer the tapestry of human civilization got. “Culture”, which we struggle to see without an object of comparison, became visible, illuminated by stories which acted like mirrors, bouncing their reflections around the world. And as the number of reflections grew, the colors in the tapestry grew richer.
Today, our understanding of our world is a direct result of what we see in that tapestry, and the stories have expanded beyond simply those that could be told through books and scrolls to a multitude of media, from photography to film to digital art. And from the very beginning of this process, right up until about 2014 or so, there was more or less a direct correlation between the volume of media produced and the depth of human understanding. The more media there was, the more we knew about ourselves, the richer the human experience.
But we have changed course. Ignorant of the qualities which made such information valuable, guided mostly by cave-man instinct, we decided that if information good and more information better, then me want make much information possible. And information, it turns out, is not unique among the things humans have sought to mass produce.
If you’re interested in food, for example, you likely noticed that over the last decade or so, there’s been a movement to bring back old varieties of plants, and to invent new varieties so as to reintroduce the diversity that was lost when we transitioned the bulk of our food system from small local farms to the industrial supply chain. You may have discovered that carrots weren’t always orange and broccoli wasn’t always green. And if you went down the rabbit hole, you may have learned that there are entire species of plants we simply don’t even eat anymore, merely for the fact that their natural qualities make them difficult to mass produce. You may have discovered that centuries ago, our gardens were filled with plants like skirret and orache and rapunzel that have, thanks to their various idiosyncratic qualities, become so obscure that people who hear the word “Rapunzel” think not of the plant, but of the girl in the fairy tale who, as it turns out, was named after the plant, which was so popular at the time the story was written. You may have had experiences like I have where, up here in the Pacific Northwest, there are plants like thimbleberry, which doesn’t appear on grocery shelves simply because it becomes too soft to be appealing shortly after being picked. And you may have learned that the grains we eat suffered a similar fate, being whittled down from thousands of varieties to about 3, because the machines we use to process wheat require that the wheat berries be standardized and uniform in size.
Mass food production was a devil’s bargain that gave us much more food, but in doing so reduced food as culture into food as calories. It took from us the quality and variety of the food. It made our food more vulnerable to disease. It corroded food culture by eliminating local and traditional species and the seasons and traditions associated with them. And now, we’re set to do the very same thing to information.
Mass production molds, manipulates, mutilates, its inputs until they fit in a box the right shape and size for the machines, separating them from their intrinsic properties and severing their connection to the web of life.
It has been 10,000 years since the first humans developed a writing system. For all but the last few decades of that, words and letters were the exclusive property of humans. That meant that any time words and letters appeared in an organized way, a human chose to put them there. No animal could use a pen. No natural phenomena could organize letters into words. And so even the simplest, most benign arrangement of letters could carry immense meaning. Like a petroglyph, we could reflect on why a particular set of words appeared where they did. We could wonder about the person who created them, and who used the tool that brought them into the world. We could ask ourselves who that person was and what their life was like, and what motivated them to think the thoughts that eventually led them to write such words. Words did not even need to be true for them to expand our knowledge of humanity, because true or not, we could be sure that a human chose to write them, and choosing to write a lie can give us as much information about the author as the truth. Why did that person lie? What was their incentive to do so? To answer these questions is to learn something about humanity far deeper than whatever untruth was written. Therefore until recently, it was impossible for words to be entirely void of meaning. There was poetry in everything.
Only in the 21st century have we begun to see meaningful separation of words from human intention. Prior to the screen age, media could be spliced and facsimiled, but even those copies needed to be directed by human hand--a human who would have known the number of copies and the cost of the paper they were printed on. Of course, we had empty clones—forms and worksheets and handouts whose spirit had been diminished through absentminded duplication—but even then, there was real information contained within them. The absentminded duplication was being carried out, somewhere, by an absentminded human whose absentmindedness was itself an information-containing artifact. And the words, irrelevant as they may have become, could still be read with intent to understand their author at the time that they were written.
It was only with the arrival of computers that we begun to see this fundamental truth change, as words were separated from human motivations. Screens meant letters no longer needed to be written, and software meant they no longer needed to be selected. And yes, screens existed long before computers, with words and images being ubiquitous on television, but television text, though it did not require a human hand to physically write it, still needed to be written in the sense that the words themselves still needed to be selected by a human. It was only with the advent of software that things started to change. A simple piece of software could simply pick words from the dictionary at random and project them on a wall for all to see, and no meaning at all could be drawn from these words. They were simply words.
This marked the beginning of the detachment of media from meaning. But even then, there was so little utility in such behavior beyond some kind of design student art project that it did not meaningfully impact our relationship with words. Deterministic software (the kind that fails to pass the Turing Test, which was all we had up until about 2022), when tasked with creating “media,”produced output that had little utility, was easy to identify as computer generated, and was for the most part boring. The real change arrived with AI.
In the early days of AI, the initial backlash centered around the fact that AI “hallucinated”, or that it was a sycophant that would lie to you to make you happy. The initial manifestations of this were so trivial they could barely be considered lies. The fact that ChatGPT would confidently tell you that 2+2=5 was less a lie than it was a “bug” in the traditional sense. But as AI rapidly improved, the mistakes suddenly become significantly harder to identify as such. Those who used the internet without their whits about them found that all it took was a brief moment of weakness to find they’d accidentally internalized a lie Claude told them with the gusto of someone who truly believes what they say. Suddenly, AI’s “lying” became the central concern about these systems, and correcting those lies the solution that would (we hoped) make everything go back to normal. But this is where we got it wrong.
While we correctly identified that AI could not be trusted, and while we were right to be hesitant to rely on the information it gave us, the reason we can’t trust it actually has nothing to do with its lies. Lying, after all, is a thing all humans do too, without disastrous consequences to our ability to function as a society. What makes AI lying problematic is the same thing that makes all AI communication problematic: that it is entirely disconnected from the incentive structure that comes with being a human in the world. It has no experience. Its words, true or false, do not mean anything.
A lie that comes from a human may be factually untrue, but the fact that a human chose to tell it still carries meaning. Whether out of fear, or malice, or prejudice, or selfishness, or cunning, or simply as a matter of culture and tradition, humans choose to lie, and the choice is a reflection of the truth. As recipients of these lies, we come to understand their authors. In knowing someone has lied to us, we know more about the world, not less. And the rare case where this is not true, the pathological case where a person lies for no reason at all, is the case of the psychopath. A psychopath is a chaos agents. Humans almost universally find these types of people creepy and uncomfortable, because we innately see their behavior as destructive to the functioning of human society.
Civilization is not threatened by the existence of psychopaths, however, because they don’t exist in numbers large enough to disrupt our societal structure. But what if instead of psychopaths occasionally appearing as individuals within our society, we decided to elevate all psychopaths to positions of authority? What if all the information we ever had—the information on which we’ve relied for thousands of years to expand our understanding of humanity—all the inherited wisdom, all the contents of our books, all the advice of our elders—what if access to all of that suddenly had to be brokered by a council made up entirely of psychopaths? And what if that information was not only being brokered by psychopaths, but written by them? And what if, when asking about information in a book, you couldn’t know whether the information you received was verbatim from the book you asked about, or merely paraphrased, or reworded, manipulated, or perhaps from a different book entirely? Perhaps even made up on the spot?
Lying was never the problem. The problem is the complete and total absence of fundamental truth—from facts and falsehoods alike.
For the first time since the invention of media itself, a piece of media’s existence alone cannot be presumed to be inherently meaningful. Text, images, music, video, are now brought into existence in a perfect type of randomness we’ve never dealt with before. Even things we previously called “random” were anchored to our reality in a way that this new technology is not. A tree that grows “randomly” on a hillside ends up where it does as a function of the same weather patterns we experience—the same rainwater we drink, the same air we breath, the same pollen that makes us sneeze. As cheesy as it sounds, everything was indeed connected. Even a dice roll begins with a human hand. AI randomness, however, cleanly severs this connection. The content AI produces emerges from the type of perfect, black-box randomness that was previously only theoretical. And because AI now functions not only to generate the content itself, but as the strategist, choosing what to make and when to make it, the value of any sort of content that AI is capable of producing, whether or not AI actually produced it—goes to zero, because not only are we inundated with slop, we have also lost the ability to know what’s slop and what isn’t.
In AI-enthusiast communities you’ll often see comments that dismiss this problem, claiming that it’s the ideas that matter, not the author’s choice of whether or not to use AI to develop and express them. Venkatesh Rao recently posted a note parodying the five stages of grief, which started with denial (“Nobody worth talking to would use AI to write”) and ended with acceptance (“The honest version is: It’s not about using AI, it’s about good vs bad ideas.”). But this is false reassurance based on a failure to grok the crumbling foundation of the old world order.
Ideas, save for the simplest and most obvious of them, are not simply stated directly, and in plain terms. New ideas, complex ideas, ideas about messy things like human culture and behavior are difficult to express. Worthy attempts to do so often depend on the messy complexity of language itself to communicate what cannot be said directly. The form of such ideas emerges indirectly, between the written lines, first as insight, and eventually as understanding. When we talk about complex things like cultural movements, for example, it’s easy to point to simple, discrete things that are commonly associated with a culture— Punks have mohawks and tech workers like Sweet Green and Gen Z has underdeveloped social skills. But none of these things are the same as understanding that culture. Understanding is something bigger than just the words on the page, and therefore to achieve understanding we must commit ourselves not just to reading literal words, but to seeing poetry in them. The poetry is the understanding. But in a world where we can no longer actually know whether written words even have insight in them to be gleaned, suddenly this analytical process becomes dangerous. Suddenly, the challenge in encountering a confusing and incoherent piece of media is no longer about seeking the truth in those words, but deciding whether or not a truth even exists to be sought. And when the answer to that question can never be known for sure, when it’s always possible that the writing you’re devoting brainpower to understanding may not actually contain anything to understand, the rational action is not to engage at all.
Where once everything was meaningful until proven otherwise, we now have the opposite: everything is meaningless until proven meaningful. The only problem is, despite what Pangram would like you to believe, this is and will forever be impossible to prove.
What does this mean for us, whose entire civilization was (and for moment, still is) built on the presumption of profound meaning in media? Our ability to collaborate, interact with, depend on—to understand groups of people we’ve never physically met, is entirely contingent on information being a meaningful product of tangible human effort. But the rug has been pulled out from under us, and Baudrillardian symbols are all that’s left. We are Wile E. Coyote having run off the cliff, but having not yet looked down. We will probably continue on like this for a short while, behaving as if media still has meaning, but it will not hold. The foundation supporting our entire societal structure has liquified, and everything built on it is crumbling. The Dead Internet is nearly manifest, and the physical world isn’t far behind. That civilizational horizon we spent thousands of years broadening is retreating all the way back to where it began: Darkness everywhere, save for the few things close enough to be illuminated by the OLED displays.
What matters in a world where media caries no meaning? Where the means of existence—food, water, shelter—are abundant, but the reason for existence--human culture, and the collective pursuit of truth—are no longer available to us?
The hall of mirrors
The AI age does not work like the one we grew up in; It is a hall of mirrors. The world we knew was one in which our creations were inherently meaningful, and could therefore be an end of themselves. The economic apparatus of society was merely the means by which we produced those meaningful things. But the AI age turns this on its head. With the creations emptied of meaning, they become mere reflections of once-meaningful symbols. The number of mirrors is ever-increasing, creating infinite reflections from a dwindling pool of source material, simulating what was once rooted and whole, ushering in the most sophisticated version yet of what Baudrillard called “Hyperreality”, where the patterns and symbols being sourced from the many ages of profound meaning that came before this one now float arbitrarily through thoughtspace, empty of all profundity, referring only to each other, and disintegrating further with each new reflection. Some of them—those whose initial images were particularly clear and bright—are still vaguely comprehensible, but they do not serve the same purpose they did before. They are no longer there to serve humanity; they’re there because the machine has discovered it can use them to control us for its own ends. Our most beloved notions—people, culture, travel, adventure, destinations—no longer tell us anything true about ourselves. Instead of humans using the machine to produce humanity, the machine is now using humanity to subjugate humans.
You can vote in the frothing nativists or the smarmy centrists or the brain-damaged insurgents of the radical left; it doesn’t matter; you will always be governed by Claude.
Once upon a time, the skeptics who worried about the power technology had over them could seek refuge in the simply avoiding the screen. When the algorithms started to dominate, about 15 years or so ago now, those with a keen sense of where things were headed could simply buy a flip phone, relegate the laptop to work tasks, and carry on with the satisfaction of knowing that the Facebook feeds that were making everyone around them so angry couldn’t touch them. But screens, in an age in which we talk to LLMs on our phones as if they were people, are but one of an ever-growing number of vectors. Even the ideas that come directly out of the mouths of people right in front of us are themselves increasingly a product of the world as transmitted to them by Claude. Our vocabulary is changing. We use words like “surfaces” now, because Claude says them to us. The “NPC” phenomenon, where peoples opinions are not their own, but easily traceable back to the TV shows they watch, is amplified one hundredfold by the fact that the content of the shows themselves are now traceable back to Claude.
In computer science we’ve long used the “Turing test” to determine whether a machine passes as a human. Now we’ve got the opposite—a reverse Turing test, to determine whether a human passes as a machine. A human whose autonomy and agency has been swapped for a stream of instructions from Claude passes the reverse Turing test; they are indistinguishable from a robot, meaning slop can now be delivered not just by text and video and audio—it can also come directly to you from another real life human.
Some people just don’t care
That state of perfect distraction is the final capitalist form. It’s what happens in a perfect vacuum—in the total absence of any and all countervailing religion or ideology, when capital is given full freedom to colonize every aspect of society. Capitalism is allergic to complexity, because complexity is not predictable and therefore can’t be monetized. It is a terminator robot for the parts of our society which rely on complexity to function, breaking apart the organic structures we see as culture in pursuit of perfect atomization, where monetized transactions can be inserted into each of the the gaps it cleaves between the nodes. Humans, then, simply become components—acting as Task Rabbits for Claude, their interests and whims and desires both resulting from, and governed by, statistical models that keep the system running smoothly. We become Truman on his show. We become farm animals out to pasture, encouraged to feel free and in charge within a closed system in which everything is accounted for, too distracted to wonder what lies beyond the fence.
Only a fool claims to know what the future holds, but it’s hard to ignore the sense that something new has already begun to take shape. Raw rationality remains the dominant driving force in our society, but where historically the only alternative seemed to be to what the trads are doing—seeking Spengler-esque shelter in 21st century style faith—we are now beginning to catch glimpses of a genuinely new tendency. The quitters and skeptics are proliferating. Chad Whitacre of Open Path, a respected software veteran and open-source champion, wrote a final blog post back in May, announcing his retirement from the entire internet and his intent to live a “neo-Amish” life rooted “not in 1780, but 1980”. Mrinank Sharma left Anthropic to write poetry. They may not yet know which path they’re walking but they’ve gleaned that this LED floodlit 12-lane hyper-rational superhighway doesn’t seem to go anywhere—so what exactly is the point of making faster cars?
It’s here that the rift is opening. Artist vs philistine. Romantic vs utilitarian. Idealist vs opportunist. There have always been those that see the world as a computable entity—a playable, winnable game—and those who see it as life all the way down—a fractal process, without beginning or end, and with nowhere to go but here. Anyone who has ever visited an art museum with friends has surely noticed these two tendencies. Some tune into the central plot line of humanity to connect with the fruits of human thought going back thousands of years and to follow its evolution over time through to today, and to understand the effort to see fundamental truths about ourselves through this process as the central work of humanity. Others, however are more interested in practical matters. They stand in front of the Jackson Pollack paintings and wonder about auction prices and debate whether or not, “My toddler could have made that.”
These two types of people have always lived amongst and benefited from each other. People who have never once wondered who built Central Park or why the buildings in Paris look the way they do nevertheless feel their effects, and the people who do wonder about those things are in many cases only able to do so thanks to thanks to the people who spend their time on more practical things. But all of us got to live in the same world of distinct beauty and personality together. The physical fabric of our places—from landscaping to buildings to infrastructure, bridges, viaducts, trains, power plants—was a contribution to our collective story, with the art of those things being inseparable from their utility. And the same was true for our media, which was held to a high standard and was not yet a synonym for entertainment. Books and films, by default, were presumed to aspire to something lasting and meaningful and to challenge us. We praised books that respected their audience, and in turn, asked something of them. Media that failed to do was seen as vice—low-status guilty pleasures that acquired derogatory labels like “pulp”—suggesting that the stories themselves were as disposable as the disposable paper they were printed on. People left theaters feeling the full range of human emotion—confusion, inspiration, anxiety, curiosity, anger, generosity. “Culture”, though not always shaped by everyone, was experienced collectively. We were all on the same boat, so to speak, even if only some of us were paddling. But as all media has degraded into hyper-personalized entertainment pulp, and the pulp is increasingly made from slop, our boat has broken apart into shards being used as crude personal flotation devices. We no longer anchor in the same shared reality, we just drift further and further apart. And so for many, the slopification of our world goes largely unnoticed. The carefully bound book jackets of yesterday give way to the YouTube thumbnail of a gaping-mouth guy saying “I can’t believe this worked!!”. The buildings that fill our cities are no longer intentionally designed works of art, but merely blocks of housing units (because everything is a “unit” now), made according to the same recycled slop plans, from the same mass-produced slop materials. We don’t have the time to opine about things like beauty and meaning. How do you measure beauty? How do you even define beauty?
Those who ignore or remain oblivious to these changes become living vectors of the machine, doing its will beyond the screen, making it ever more difficult for the artists (a term I’ve begun to use for those who wish to remain beyond the machine’s reach) to persist. The artists wish to preserve their capacity for original thought outside the mechanism of economic efficiency, and avoidance of screens is no longer the shield it once was. The new task becomes reducing exposure to entire groups of people, who come to be seen as impure and corrupting—people who have made something else their god.
The artists will develop a mechanism for identifying and protecting each other rooted in a deep, Amish-like skepticism of the out-group. And like the Amish, it will be this skepticism that in turn makes them valuable to society, for from the margins, they will be among the few remaining sources of genuine novelty. Their work products will be prized not just for their function, but for the value system that produced them, and their work will be imbued with a special quality which relies on the devotion to an increasingly religious code of conduct that guarantees values compliance in a way that legal law cannot. This cultural ruleset will be what separates the true artists from their machine imitations.
And as it always does, the machine will hijack and exploit the fruits of those artists’ work. It will convert truth into detached, mass-produced symbols, and sell them back to us. It will generate imitators who attempt to codify a secular stand-in for culture. It will make rules about how and when various tools can and cannot be used, and to what extent. It will use legal language to quantify limits and exceptions. It will have rulebooks full of phrases like, “reasonable effort” and “responsible use”. It will set up systems like Pangram. And ultimately, it will fail miserably in the same way that all others who have attempted to systematize complexity have. Cheating and performance enhancement in sports, corrupted “organic” food and farming—writing concrete rules to govern these system will simply incentivize people to find the loopholes. The spirit of the law is lost to mere technical compliance. The word “organic”, which was originally meant to communicate a meaningful relationship to food, and soil, and ecosystems, has been degraded into a empty marketing term, hijacked by the machine and used to increase prices, and the farmers producing the type of food originally meant to be called “organic” are ironically no longer able to calls themselves such. In the end, we end up right back where we started: the only way to validate the food you buy is to know the belief systems of the people who grew it. And the only way to validate the ideas you consume is to know the belief systems—the culture—from whence they came.
Only culture, not law, is capable of legislating complexity. Culture is the antidote to unsavory rational incentives, the goo that fills in the cracks and loopholes and subjective interpretations. Properly established culture is far more powerful than law, because scorn, and shame, and pride can regulate behavior that black and white law cannot.
Culture is the reason that, in the game of Craps, players to play the don’t pass line, despite it being entirely legal and having the best mathematical odds. We simply do not do that.
Culture is the reason that fights break out in hockey games, despite the automatic penalty and zero rational incentive to start them. It’s simply what we do.
Culture is the reason that in China, despite it legally being a day off, workers are expected to be in the office on Saturday to “get extra work done”, but the same expectation does not apply to Sunday, which is a day of rest whose sanctity shall not be violated. Yes, we’d get more work done being there on Sunday too, but we simply do not do that.
Culture is the reason that, though all countries have laws against thievery, European bicycles regularly have their locks cut whereas in Japan, there very concept of even using a bike lock is nearly unheard of.
And culture, at the end of the day, is inseparable from religion. The aversion to theft in Japan is inextricably linked to the religious idea that lost objects are the work of spirits, and that engaging with them will bring misfortune. It matters little whether individual Japanese people actually believe this to be true—what matters is that they can all assume everyone else is aware of this collective cultural belief. Instead of the western thought process, which says that, “If I don’t take it, someone else will,” theirs says, “I won’t take it, because no one else is going to either.”
The anti-AI religion may not begin as a religion of gods and deities—those stories will emerge over time via the adoption of stabilizing but rationally inexplicable behaviors. Christianity identified the seven deadly sins—pride, envy, wrath, gluttony, lust, sloth, and greed—as destabilizing vices to be contained. Buddhists, in the same way, are deeply skeptical of materialism. This new religion will do precisely the same thing with AI, and it will manage its members relationship to technology in the exact same way: not with law, but via the tools of culture: guilt, shame, pride. Whether or not certain behaviors are “technically legal” or “not explicitly prohibited” becomes irrelevant. There is no line between what constitutes gluttony and, “just eating a bit more than I usually do”, between sloth and, “just being a bit tired lately”, between greed and, “fiscal responsibility”. It matters little whether someone who is perceived to be greedy is “actually greedy”. And the Amish do not have hard rules about the amount of time its people are permitted to spend interacting with the “English” (as they call us non-Amish). All that matters is how the group perceives its members’ behavior—whether you’ve got the gestalt of someone who has fallen out of line. Anyone who hopes to remain part of a community must actively work to protect the reputation that rational incentives never stop threatening to undermine.
The new religion will function in the same way. AI judgement will be less about whether or not someone used it for this or that, but simply about whether he is seen to be “associating too closely” with it. And via this mechanism, a new culture, and a new religion will grow, take shape, win converts, and become a force to contend with in the coming years.
Either that, or it won’t, and we’ll all be swallowed up by the machine.
. . .