Their gospel is to “trust the system”. Their sacrament is the model. Their blessing is “data-driven decision-making”.

The New Priesthood of Code and Smoke

“Drink the long draught for the Big Priest”

The Fall, New Big Prinz

“And Aaron said unto them, Break off the golden earrings, which are in the ears of your wives, of your sons, and of your daughters, and bring them unto me. And all the people brake off the golden earrings which were in their ears, and brought them unto Aaron. And he received them at their hand, and fashioned it with a graving tool, after he had made it a molten calf: and they said, These be thy gods, O Israel, which brought thee up out of the land of Egypt.”

Exodus 32: 2–4

Once, we feared the gods. Then we feared the markets. Now, we fear the machine.

In each era, we invent a system bigger than us, a system that is unknowable, self-justifying, and we let it rule us. Today, AI has become that system. We call it the machine. We do not only use it, we increasingly defer to it. We trust it to guide hiring, to shape content, to determine credit scores and even to flag criminal risk. More often than not, we do not even know how the decision was made.

Thus, a new priesthood is born. These are the AI whisperers. The interpreters of code. Engineers, executives, consultants form the new clergy, shaping reality on our behalf. Their gospel is to “trust the system”. Their sacrament is the model. Their blessing is “data-driven decision-making”.

Whole organisations now make strategic moves not based on human intuition or lived knowledge, but on what they think AI might soon be able to do. It is all about not what is, per se, but what it promises; what it could be. The singularity is always just around the corner. Thus, investments pour in based on speculation and not demonstration. And in the meantime, we build entire temples around the prophecy.

We have gone from providing steer, to oversight and now to worship.

Magical Thinking in a Lab Coat

We have always wanted a magic pill, a silver bullet, a fix-all. We have always needed salvation. Many hoped that science would bring it. Others put their faith in the metaphysical, the supernatural. Many of us are pinning our hopes on AI.

We believe it will save us from inefficiency, from disease, from collapse, from the cavalry charge of the polycrises. We believe it will spot a threat before it emerges, balance the system that we cannot, solve the problem we have spent decades failing to fix, or kicking the can down the road for future generations to resolve.

This is magical thinking: faith in the form of code.

What if this “magic” is just a cover for cowardice? For abdication? A way to outsource hard questions? A fantasy that lets us keep doing what we are doing, regardless of the potential damage, because we believe that something smarter than us is working on the fix?

When we say we want AI to help, what we really want is for AI to save the day, without requiring anything of us.

And yet, magical thinking rarely ends well. In myth and literature, the magical solution often comes with unintended consequences. The wish granted is never free. The monkey’s paw, the pact with the devil. It all comes due.

The same lesson applies today: every miracle promised by AI demands a cost. If we ignore the price, we do so at our peril.

The Abdication Economy

The most dangerous thing about AI is not that it will replace us but that it gives us the perfect reason to be cruel and, at the same time, back away from accountability. It gives us an excuse to stop thinking critically, to click “accept” on outputs that we do not understand, because they are clean, fast and algorithmically blessed.

It is about ethics, creativity, decision-making and governance. It goes far beyond just being about jobs. We are building a world where the first coat of paint comes from the machine and the second, what used to be human craft, is deemed a nice-to-have.

Why pay a translator who wrestles with nuance when the post-edited AI version is “just so”? Why commission a writer who pushes your thinking or even blows your mind with a certain analogy, metaphor or turn of phrase, when the prompt-fed model returns something “good enough” in 0.5 seconds?

This is erosion. The more we accept it, the more we convince ourselves that the surface is enough.

And here, we need to speak plainly about the growing culture of abdication, particularly among the architects of AI itself. When the Big Priest Sam Altman says “We didn’t expect this to happen”, it raises the question: What, precisely, were they expecting? Whether this is sincere concern or strategic deflection, the effect is the same: the creators position themselves as both prophets and bystanders, warning us of dangers while continuing to accelerate them. They cannot be both and we all need to wise up to this culture of abdication.

So, they warn of the obsolescence of the knowledge worker, a fear now echoed across media. Yet few warned with such urgency when manual workers were displaced by automation. Why now? Is the threat real or is it manufactured to keep us on edge, frightened and obedient while we await the next announcement from the high priests of the Valley?

It is the Frankenstein defence in a modern form. The monster has escaped, the lab doors are open, but the creator shrugs and says, “We had no idea this would happen.” And so we return again to abdication, not just of responsibility, but of foresight, control and moral leadership.

We should be sceptical when warnings come from those who profit most from the thing they warn against. It is like they hide in plain sight. Is it genuine concern or clever branding? Are we being alerted to danger or herded towards a predetermined path, one that benefits the AI clergy above all?

In H.G. Wells’ The Island of Doctor Moreau, the question of creation, control and cruelty loom large. Are creators responsible for their creations? How do we treat entities with semi-autonomous capabilities? Moreau’s beast-men echo today’s autonomous systems: imperfect, confused, caught between command and autonomy.

Wells introduces us to the House of Pain. This is a place of transformation, obedience and suffering masquerading as progress. The law was simple and cruel:

“He who breaks the Law goes back to the House of Pain”

“The Sayer of the Law: His is the hand that makes. His is the hand that hurts. His is the hand that heals. His is the House of Pain”

What is the House of Pain in today’s tech world? It is abstraction. It is the threat of pain, the fear of pain and the pain itself. It is the denial of recourse when the system fails. The silent despair of being flagged, filtered or fired by forces you cannot see, for reasons no one will explain. The House of Pain is no longer physical. It is in our heads. It is coded into the system, distributed across platforms and wrapped clean in UX.

And just as Moreau’s creations feared breaking the law and facing retribution, so too do we exist under constant, ambient threat. All of the time. Not of a whip of a cage, but of cancellation, deplatforming, deactivation or silent algorithmic exile. The pain is quieter now, but it is internalised. And it is always there.

This Is the Hand That Uses the Tool

The fantasy of AI-as-scapegoat allows us to skirt around one core truth: technology is not neutral. Tools reflect the intent of their users. And when we see AI as a black box of destiny, we absolve ourselves of the need to examine our own intent.

We have to stop asking what AI will do to us. We have to start asking what we are doing with AI and why.

This is essentially a conversation about human choices. AI did not build our economic system. AI did not decide profit mattered more than people. AI did not cut corners. We did all of that. AI is just the extension of a mindset we have not questioned and tolerated for too long. This is a mindset that confuses scale with success and detachment with objectivity.

We need a culture of responsibility that starts long before a product ships. Design must include dissent. Deployment must include accountability. And when things go wrong, as they inevitably do, we need people and not policies. We need them to step forward and say: “We will fix it.”

From Obedience to Ownership

There is a way forward. But it starts with reclaiming ownership. AI is not really the problem. AI is mindless. It is our belief in its authority that is the problem. It’s the story that we tell about AI: that it is smarter than us, that it is more neutral than us and more capable of sorting out our mess. This is the fiction.

We, the humans, need to be in control, not only of the narrative but how we handle AI in real life. We have to stay at the helm. We have to stay the master. We have to be rigorous, transparent, accountable, interrogate every output and every decision downstream of it. Train not just the model but the people who use it. Teach ethics together with AI engineering. Build policies that force human-in-the-loop and not human-out-of-the-blame.

The minute we stop steering, we stop being responsible.

The minute we stop being responsible, the system does not belong to us anymore.

We are not helpless in the face of innovation, but we must choose to be active participants. That means asking better questions, demanding better answers and resist the pull of easy automation when it comes at too high a cost.

The Ghost Was Always Us

The great fear is not that AI will become sentient but that we will stop caring whether it does because we have already handed over our autonomy.

It is not too late. We can choose to use AI as a tool, not a crutch. A partner not a priest. We can resist the seduction of magical thinking and look, instead, at the reality in front of us: a tool that reflects what we put in.

The ghost was always us, not the machine.

We Will Need a Doctrine

We may well dismantle the gods, but we do not abandon worship. In the vacuum left by disenchanted mythologies, we seek new symbols, new sacred geographies. So where is the new Holy Land? Silicon Valley? A data centre cooled by glacial meltwater? The pristine server farms in remote deserts were the servers hum and pulse like a techno-cathedrals? Should we build new Nazca lines there?

And what of the New Jerusalem? Is it the utopia promised by singularity evangelists? A world of post-scarcity, eternal health and digital transcendence? A place where minds are uploaded, labour is obsolete and the code runs flawlessly?

What of the afterlife? The cloud? Is it consciousness as software, liberated from the burden of biology? Or simply digital relevance, a life extended by metrics, influence and algorithmic memory?

And if we do achieve this imagined reward, what is the price of entry?

What is the new moral code? It is not written in commandments or carved in stone. It is etched in terms of service, enforced by moderation algorithms and preached through TED talks and press releases. The code is: Be Innovative. Be productive. Be data-driven. Be optimisable. This is the most important part of the code.

And what is the reward for obedience? A frictionless life. Personalisation. Efficiency. Convenience. Survival in the age of digital acceleration. A seat at the table of relevance.

What will be the new music of worship of this age? Will the hum of data centres become the drone of future contemplation? Will the quiet pulse of cooling fans and server racks inspire a future Bach, as the echo of cathedral vaults once did? Just as Gregorian chant emerged from the acoustics of stone chambers and the clang of heavy industry shaped 20th century riff, what new creativity will be born from this synthetic landscape?

Will we hear the sacrosanct in the algorithmic? Will a future bend strings to the dissonance of AI hallucination and cybersecurity alerts? Will we compose elegies for lost data or anthems for uploads that never complete? Is it possible to imagine such horror?

And what of visual art? Will we build new triptychs, not in oils but in code? Will we raise spires made of carbon fibre and steel, stained-glass windows rended in OLED, temples that transmit prayer via Bluetooth, churches tracked with QR codes?

Will this techno-faith be universal or culturally fractured? Will some worship the code in English, others in Mandarin, other in silicon dialects not yet born? And where there is belief, will there be dissent? Will there be heretics who reject the machine, who question the logic of optimisation? Will there be schisms between organic and synthetic, human and post-human? And if so, what will be the penalties? Social exclusion? Algorithmic ghosting? Biometric exile?

Where will transhumanism sit in this theological architecture? As sainthood? As sacrilege? Will the augmented be revered as pioneers or feared as apostates?

The sacred always comes with sacrifice. The question is: what are we sacrificing to reach this digital promised land? What are we leaving behind? Will the god we meet at the top of this mountain still recognise us as human?

If the ghost was always us, so too is the hunger, for coherence, for clarity and for belief. This is where we reach the limits of our old stories. The scriptures of ancient faiths say little — actually nothing — about algorithmic fate or the metaphysics of machine intelligence. It is not that sacred texts are incompatible with techno-futures, they simply lack the lexicon. To keep them relevant, we must bend their metaphors far out of shape, reinterpreting prophecy through silicon and spirit alike. We have already been doing this for some time.

What would be the new scripture? What would be our Book of the Machine? Would it be open-source? Would it be versioned and forked, annotated by engineers and ethicists alike? Would it be immutable like stone tablets or ever-shifting like a blockchain ledger? Would its prophets be founders and futurists, its parables delivered in keynote speeches, its revelations encrypted in code?

And how would we test its truth? Through computation? Through consensus? Through divine signal emerging from the noise of endless data?

We are writing our new mythos as we build it, line by line, prompt by prompt. The question is not whether it will have scripture but if it will have soul. If we dismantle the altar of AI, something else will take its place. Humans are myth-making creatures. As Jeanette Winterson wrote: we need a new kind of humanism, that incorporates the machine without becoming the machine.

That is the work ahead, It is not just about resisting AI’s mythos but writing a new one, something post-human, transhuman, spiritually mature. A worldview where AI is neither god nor golem, but a collaborator. A belief system that anchors us not in obedience to black boxes but in responsibility, humility and expanded imagination.

This article began a couple of days before the launch of ChatGPT-5 and was finished — up until this very word — a few days after it hit the world.

Can we be honest and say it was pretty much more of the same? A let-down is not quite the word, but feels like it marks something, a harbinger of something that we cannot quite put our fingers on.

It is a let-down because it is showing the frayed seams of the hype arc. Or is it a hype circle? It feels less like a revelation and more like the moment the music falters at the party or in a bar, when you realise that maybe this is as good as it gets and the “next big leap” is maybe just a marginal shuffle. This may be the first hint that the AI bubble may already be cresting.

We have seen this many times before: the dot-com delirium, the tulip mania, the frantic rush to hoard bronze that heralded the collapse of the Bronze Age civilisations in the Near and Middle East. The story is forever the same: the prophecy runs faster than the reality, the promises outpace the product and when the crash comes, it is never just financial.

So what of the human cost already? The job losses, for starters. The industries quietly pivoting now that they have decided they don’t need the human at all. The skills atrophying, the talent pools drained. The people told they were essentially until the spreadsheet proved otherwise.

If the bubble bursts and the AI revolution does not deliver the goods, how do corporations lure those people back? When the hype machine winds down and the zeal fades, when the venture capital dries up, what happens to the new religion, the new revolution? What happens to the fervour, the zealots, the believers who traded their craft for the promise of a frictionless machine?

If it is all just hype, if it turns out that behind the sermons and pilgrimages and the sacred UX there is nothing more than a crock of shit, what then?

After everything? After all of this? What happens to those AI overlords?

People will parade their heads on sticks.

Mark Gibson

August 2025