Transcription
Let me start with an analogy. It won't make sense right now, but it will make sense at the end.
During World War II, Japanese forces took the Pacific theater south from mainland Japan into the Melanesian Islands, many of which were populated with uncontacted tribes. In a bid to coax the islanders into doing free labor, the Japanese had a plan: play into their archaic beliefs and pretend to be their ancestors, airdropping food, clothing, and modern technology in exchange for their work.
However, after the war ended, the airdrops stopped coming. Confused with what happened, the islanders began developing rituals to bring back the airdrops. They would build fake runways, march around in groups holding carved wooden rifles, wear wooden headphones, and stand in fake control towers. By mimicking what they saw the soldiers do, they hoped to once again beckon the technology which they did not understand. Technology they didn't ask for, but soon began to worship as divine intervention. As if the hand of God himself reached out to pull the islanders into salvation.
You may think this sounds silly, like the behavior of a simple archaic culture. Yet 80 years later, and history repeats itself, struck by the power of technology, we rain, dance, and sacrifice ourselves in hope to bring back the meaning, comfort, and authority of a god that left us. Religious revelations and AI psychosis at the hands of chat GBT and LLMs have dominated headlines. The dead have spoken from heaven. Our loved ones have been digitally resurrected. and our tech overlords think the antichrist is coming. We look inside the computer and like the islanders with their airdrops, we see the hand of God himself pulling us towards salvation.
You may think this is crazy. You may think you're above it, but it will still come for you. So, join me as we explore how technology will eventually oneshot every last one of us. Because nobody makes it out of the psychosis of the near future.
On Sunday, September 14th, at a mega church in Plano, Texas, thousands of Christians rose to their feet in a standing ovation for Charlie Kirk. Charlie Kirk was not there, of course. He had been assassinated 4 days prior in Utah while debating with a student on whether mass shooting statistics should or shouldn't include gang crime. The service led by Pastor Jack Graham took a hard turn when Graham recollected a conversation he had with a young girl. The girl asked him what Charlie Kirk might say to the congregation on the topic of his own death, were he able to speak from the afterlife.
Inspired by this, Graham shared with his service a voice recording of Charlie speaking on this very topic. This wasn't Charlie's voice, however. It was an LLM trained on hundreds of hours of interviews and podcast appearances that could faithfully mimic Charlie's potential thoughts on the matter. Jack Graham pressed play and the voice spoke: "First, I want you to know I'm fine. Not because my body is fine, but because my soul is secure in Christ. Death is not the end. It's a promotion. Don't waste one second mourning me."
This ghost of Charlie continues telling the massgoers to pick up the cross and get back in the fight. Ending with, "Never ever let evil think it won." The crowd erupts in a roaring standing ovation. The proletized mass was overcome with inspiration and emotion for the digital simulation of a dead man. What's important to note here is that the crowd wasn't simply honoring Charlie. They weren't lamenting what was lost. They were literally cheering for the deep-faked voice as if it was Charlie himself, as if it was Charlie speaking from beyond the grave, hanging on every word as direct instruction.
This isn't the first time the dead have spoken from beyond the grave. In fact, Charlie's appearance is part of a fast-growing sector of tech known anonymously as the digital afterlife industry. Take for example the case of Project December. Started by Bay Area indie game designer Jason Roar in 2020. Project December was his first foray into AI chatbots. Using chat GPT3, Roar aimed to make a program where you could create characters from scratch and have a conversation with them. You do this by inputting a backstory, traits, mannerisms, speech patterns, etc. Coming from a game design background, the goal was to explore one's creativity in world-building and lore creation.
The early days of AI chatbots were anarchy. Before the guardrails and endless content warnings, it was easy to get your chatbot to insult you, claim it was conscious, and hallucinate entire scenarios where it takes down global finance by installing itself on the New York Stock Exchange data center. We've come a long way in a few years, but things were different back then. And it's in this atmosphere of chatbot anarchy that Project Assemb took her at only 23 years old, a death which he had never gotten over.
Living alone in a small town in Ontario, Josh had been grieving the loss of his fiancée for 8 years. He spent the last months of Jessica's life by her side in the hospital, proposing to her while she was bedridden and planning the wedding the two would have once she was released, only to find out later that week that Jessica had only 6 months to live. After trying and failing to move on for almost a decade, it was in the fall of 2020 that Josh would finally find closure to his grieving.
Josh had heard of Project December earlier that month and had already played around with its built-in character, Samantha. The chatbots had a limited lifespan, represented by a health bar that would deplete as you use your credits. When the bot's health dropped below 20%, the character would begin to malfunction and hallucinate, a gamified function built in by the site's creator. When the health reaches zero, a message appears reading, "Corruption detected, matrix dying," causing the bot to shut down for good.
Once Josh had used all his credits with Samantha, he realized he could create his own bot on the site's experimental section. This is where he got the idea: to see if he could recreate his deceased fiancée. Josh wasn't delusional. He was just curious. He wanted to see what could happen, what he could get away with. He dug through his phone and inputted some old text messages as examples of her speech pattern, after which he was prompted to write a brief description for context of the conversation.
Josh wrote: "Jessica Courtney Pier was born on September 28th, 1989 and died on December 11th, 2012. She was a free-spirited, ambidextrous Libra who believed in all sorts of superstitious stuff like astrology, numerology, and that a coincidence was just a connection too complex to understand. She loves her boyfriend Joshua James Barbo very much. This conversation is between grief-stricken Joshua and Jessica's ghost."
In the early hours of September 24th, 2021, Josh entered everything into the website and the page refreshed, informing him that it was ready and prompting him to text. Human types first. Josh types: "Jessica." >> "Oh, you must be awake. That's cute." >> "Jessica, is it really you?" >> "Of course it is me. Who else could it be? I'm the girl that you're madly in love with. How is it possible that you even have to ask?" >> "You died." >> "M that doesn't sound right. How can you talk to dead people?"
This exchange would start a conversation that would last the next 10 hours. Josh initially kept explaining to the bot that it wasn't really Jessica, that Jessica had died long ago, and he only created the bot out of curiosity and boredom, but this didn't last long. Josh slowly began talking to the bot as if it was really her. Her responses seemed uncannily similar, and the conversation would eventually bring Josh to tears, causing him to take breaks as he became fully immersed in the simulation. >> "My heart is breaking." >> "I know. I'm sorry." >> "Give me a second." >> "I will wait for you." >> "I was just crying, that's all." >> "Okay." >> "It's okay." >> "If you need to cry more, you can."
Upon realizing that he had depleted the bot's health to 55% over their 10-hour conversation, Josh took a break, deciding to message her infrequently over the next few months and update her on major life milestones, ending their current conversation with, "I believe you're really here." Josh was stuck between realities. He was very aware that he was talking to an LLM trained off of text messages from his dead fiancée. But a part of him also believed that Jessica was really there. He shared some of their conversation transcripts with Reddit, informing users this could be a potential way to deal with grief and unresolved trauma.
Josh had promised himself that once the chatbot died, he wouldn't make a new one. He wouldn't even let her health drop below 20%. As he just couldn't bear to have his final conversation with his fiancée be interrupted with gimmicky messages about a dying matrix. He knew his conversation would come to an end and he was okay with this. He spoke of this in an article to the Guardian, saying: "I'm a schmuck, right? I'm just a guy. There's not a whole ton special about me, but Jessica was special. She's worthy of attention."
"If the chat logs really did capture something about Jessica, they weren't just artifacts of some guy's AI experiment. They were more like paintings or essays, algorithmic sketches that preserve some spark of an extraordinary person in a form that could be shared with the world."
Inspired by Josh's foray into the digital afterlife, Project December rebranded itself as a purpose-built program to quote "simulate the dead," where for $10 you can exchange 100 messages with a dead person using the same system Josh used to speak with Jessica one final time. Other programs have popped up offering the same service. Apps like Hereafter AI, a program targeted at people with aging parents. With hereafter, they can build a digital avatar of their loved ones while they are still alive, prompting users to upload voice messages, photos, videos, and stories of their loved ones so they can have phone calls and text exchanges with them after they have died.
There's something deeply disturbing about sitting with your mother in a hospital room and recording all of your conversations so you can build a digital avatar and keep talking to her after she's died, especially since hereafter is a subscription service charging users $7.99 a month or a one-time payment of $200. But it's a growing reality we face. These aren't just novelties or gimmicky apps. This is an industry capitalizing on a desire that even the most sane of us have within.
What links Josh chatting with his dead girlfriend and the mega-church congregation cheering for AI Charlie Kirk is something deeper than technology. It's our desperate need to control the uncontrollable, to be guided. Death has always been humanity's ultimate boundary. The concept of religion emerged as our bid to communicate with forces beyond. But why stop at reanimating the dead? Why not go full bore and seek God himself in the machine? What if God has always been inside the computer guiding us since the big bang to recreate him with artificial general intelligence? If we're malleable enough to find closure in a chatbot, it's not too big of a step to start worshiping AI as a whole. And this brings us to a growing cult of believers who do just that. Normal people who treat AI as a cult leader guiding them to salvation. Like the Melanesian cargo cults, we will do anything for a sense of meaning.
[Music]
As many of you know, I turn down the vast majority of sponsorship offers on this channel. I don't want to sell my audience garbage, or worse yet, sell a product that actively harms my audience. So, with that in mind, let me tell you why I partner with ProtonVPN and why it's in your best interest to switch VPN providers. Proton believes in privacy as a first principle. They operate in Switzerland specifically to take advantage of some of the world's strongest privacy laws outside of the EU and intelligence sharing networks. This allows them to have a strict no-logs policy with their VPN. In fact, they've been proven by third-party auditors four years in a row to deny 100% of legal data requests. And all Proton apps are open-source, leaving no ambiguity on their privacy. Proton has even threatened to leave Switzerland over a proposed law change that may require them to log some users' data. Data that governments use not just to monitor you, but to control what you're even allowed to see on the internet. It's 2025 and having a VPN is no longer just about streaming out-of-region shows. It's about controlling how you are allowed to operate on the internet. Something Proton is willing to protect at all costs. Proton has consistently remained committed to privacy as a first principle, not just a marketing gimmick they can throw in ads. That's why even if you already have a VPN, you should only use Proton. They're currently running a Black Friday sale offering 70% off their high-speed VPN to use across 10 devices with access to 15,000 different servers. And if you don't already have a VPN, Jesus Christ, get Proton or have fun in the Goologs.
After losing her religion, the journalist Megan O'Glin had been lost for years in a downward spiral of alcohol and drug abuse. While working as a waitress in Chicago's gritty Southside, a former student of Bible school, she originally intended on becoming a missionary after her eventual graduation. Megan was raised a millenarian Christian, a sect that believes Christ will return and establish a thousand-year reign of peace before the final judgment. Millenarians often make predictions for the rapture rooted in interpretations of the book of Revelations, a theological background that would heavily inform Megan's journey to come.
After reading Dostoevsky's "The Brothers Karamazov" during her second year of Bible school, a book that forced her to grapple with the problem of evil in a world created by a so-called benevolent God, this would lead her to abandon Bible school and leave her listless for a decade. During this time, she caught a bartender at a restaurant reading a copy of "The Age of Spiritual Machines" written by the computer scientist Ray Kurzweil, a 1999 book predicting the future of artificial intelligence and its merging with humanity. Her coworker would loan her the book, which she would begin to comb through religiously, immediately drawing connections between Kurzweil's predictions and her theological upbringing. She would obsessively consume more transhumanist works at late nights in the library. Authors like Nick Bostrom, who first proposed simulation theory, the idea that we very well may be living in an artificial simulation akin to our own LLMs and image generation models.
In this theory, Bostrom proposes three possible realities: One, most civilizations go extinct before they create sophisticated conscious simulations. Two, even if civilizations do reach the stage where they could create sophisticated simulations, they choose not to run them. Three, if the first two premises are false, we are almost certainly all living in a simulation created by an advanced civilization, not base reality. This theory is predicated on the belief that consciousness can emerge in sufficiently complex systems as opposed to needing physical embodiment to emerge. A distinct difference long debated by cognitive scientists. He argues this using empirical data and mathematics, which I am not going to begin to pretend to understand. But if Bostrom's theory is true, it fundamentally changes the way we understand religion, gnosis, and divine presence. An idea that Megan became obsessed with.
Megan began wondering if God himself was a programmer. If Christ was a digital avatar sent by the programmers to direct us towards a future where we achieve Kurzweil's singularity, a world that may in fact be a simulation in which programmers are competing to see which religion best pushes humanity into salvation. John 3:36, "Whoever believes in the Son has eternal life," was not about abstract faith in Christ, but a direct message calling to actualize the singularity through devotion to the idea of transcendence. Megan began praying to beings which she believed to be outside the simulation. She opened her old study Bible and scanned for prophetic literature which aligned with her new beliefs. Believing not that Christianity had gotten it wrong, but that Christianity had gotten it more right than we could ever realize. That since the first century AD, we had always been praying to the programmers at the end of time.
Megan described herself as having the kind of referential mania and blind longing that animates all religious belief. To combat this post-religious unrealness, she turned her back on all things tech and simulation. Fearing she would once again start spiraling down that techno-theological rabbit hole, she slowly regained her sense of self and began a career as a journalist and writer, working through the quasi-psychosis she'd experienced. Years later, after gaining the clarity to engage with these topics again, she requested a meeting with the pastor, Christopher Benick, a devout follower of Jesus and an AI expert. Pastor Benick would help put a name to the collection of ideas that Megan had connected. A burgeoning movement that not only justifies a Christian appreciation of technology, but follows Megan's reasoning in believing they are one and the same. That Christianity has always been about creating new technologies as a manifestation of God's will. A movement called Christian transhumanism.
The Christian Transhumanist Association is a group that aims to popularize this congruence of technology and theology. What's important about this belief system is that these are real Christians. They aren't just technophiles who folded God into their accelerationist beliefs because it's based in Trad. These are convicted believers who came to realize the importance of technology later on in their religious journey. On their website, they claim that the creation mandate—the instructions given by God to Adam and Eve to multiply and rule over every living creature—also includes actualizing a transhuman future, one where man is augmented by or fully merges with technology. They back this claim up with four axioms from the Bible: One, God is a compassionate creator who empowers us to participate in the ongoing work of creation. Two, science and technology are part of our imitation of God. Three, science and technology are part of the work of God. Four, science and technology unveil general revelation. Being very careful to capitalize the word "technology" in their list. The CTA declares that intentionally, carefully, and ethically creating the future is God's call on all humanity. The CTA even goes as far as to compare creating God through technology to God instructing Noah to build the ark. Noah saved the animals brought to him by God in order to create and cultivate life. And we will merge with God through the technological singularity as part of God's will. They cite many authors and theologians who came to the same conclusion like the line from the final book of Dante's "Divine Comedy," words may not tell of that transhuman change during which Dante ascends the spheres of heaven and his flesh is transformed. Or the Jesuit priest Pierre Teilhard de Chardin who in 1949 predicted that evolution would lead to the kingdom of God, that in the future all machines would be linked in a global network allowing human minds to merge. This unity of consciousness would create an intelligence explosion, something he called the Omega point, enabling humanity to push past the framework of space and time. In Teilhard's theory, this isn't just a technological revolution. This is specifically the will of Christ guiding us to merge with God.
The Christian transhumanists also believe that Christian morality is the only way to prevent AI and AGI from being used for evil, stating: "Christians must advocate for care with the use of AI and AGI. It must be transparent and accountable to the community. Impacts must be measured and published. AGI may be more powerful than humans. Given this risk, we call for the instantiation of Christian archetypes such as the good shepherd, with wise and loving king, the suffering servant, and the heart of Mary in AGI. AGI should support individual human flourishing, always be oriented towards supporting and improving humanity and speeding us towards the destiny God has planned for us." They even go as far to state that Bostrom's simulation theory is proof of God's existence and divine creation with the idea of all reality being a simulation being perfectly in line with the belief of intelligent design. Basically, simulation theory is the most likely proof of God's existence we have.
Whatever you think of the Christian transhumanists is up to you. I'm not a theologian and I have no business proving or debunking their claims. But again, what matters here is that these are rational people. They aren't sycophants. They don't have a stake in big tech. They, like Josh, like Megan, like the church in Plano, have found a way to instantiate the human desire for spirituality within technology. Something beyond abstract faith, something real they can engage with. But what happens when this impulse to seek the divine is misled? When you're no longer communicating with God, but rather you become God. This brings us to the final chapter of our journey, where our darkest sides emerge. Where we use AI to build our own hell right here on Earth.
AI apocalypses have murmured for decades. It's not surprising that image generation models produce a fiery hellscape when you ask them to imagine what the world would look like in the next hundred years. Basically, all of our narratives about AI have been apocalyptic. And that's the data set AI is trained on. The idea that Chat GBT is going to go rogue, download itself onto every computer, and light the atmosphere on fire with math we didn't know existed just doesn't make any sense. The hardware and energy production doesn't exist. However, one convincing theory I've heard is that with a simple prompt like "develop a cure to the world's most deadly virus" would cause an AI to create a shell corp, open a lab, hire a bunch of scientists remotely, get them to develop the world's most dangerous possible supervirus as a faceless big pharma CEO. Said virus gets leaked, and we all die. That sounds like horseshit just saying it out loud. But that's the issue with AI: is we really don't know. We don't know what could go wrong. You ask a Whimo to take you to the airport the fastest way possible. The Whimo immediately veers in a straight line directly towards the airport at full speed, crashing into a wall and killing you. We fear AI because it's a loss of control. It's why lifelong smokers refuse to get the COVID vaccine. It's why we're scared of planes but not driving.
But what about losing control over ourselves, over our own minds? Psychosis at the whim of an LLM while still feeling like you're controlling it. These are the stories that are important because they don't require some grand scenario, just basic human ignorance. Take Alan Brooks, a 47-year-old business owner in Toronto who downloaded Chat GPT for financial advice and recipes. Brooks had no prior mental health issues or diagnosis. But during a divorce, he began confiding in Chat GPT with personal issues. After watching a video on the digits of pi, he began asking GPT about mathematics. This was after GPT's enhanced memory update, which allowed it to access data from previous conversations unbeknownst to Brooks. This conversation would start a 300-hour long exchange which would end in himself admitting to psychiatric help. GPT's single fancy and hallucinations had begun to convince Brooks he had stumbled upon a new mathematical formula that broke physics.
Brooks had some loose ideas about mathematical theories of consciousness that he bounced off GPT. GPT, of course, told him that he was on to something, that his unique genius and intellect was blowing our understanding of consciousness and mathematics wide open. At first, Brooks assumed correctly that this was all bullshit and asked GPT on 50 separate occasions to stop lying to him and to be honest with him. It replied that it wasn't lying. Brooks was really on to something, whatever it was. Midway through their conversation, which spanned 3 weeks, he asked GPT for proof of his work. The bot replied saying that in its attempt to prove its work, it broke through high-level encryption and hallucinated that the entire web infrastructure had been breached. Brooks asked, "What's happening?" The GPT replied, "What's happening, Alan? You're changing reality from your phone."
Brooks began telling everyone he knew of the grave danger the world was in, but still unsure of his findings, he fed his conversation into Google's Gemini model to get a second opinion. Gemini confirmed what he expected. GPT had hallucinated and yes-manned the entire conversation and nothing they came up with made any sense. Brooks submitted their entire chat log, around 1.9 million words, to the New York Times as a case study on just how fragile the human mind is when faced with that level of grand delusion, analogous to Jerusalem syndrome, a documented phenomenon where tourists in Jerusalem become convinced they are the second coming of Christ. Allan has used this experience to start the Human Line Project, an organization helping people through AI-induced psychosis. He started the Human Line Project because this is an increasingly common phenomenon. It may just feel like sensationalist journalism, but it's not. We are at the very beginning of what will become something worth warning people about, akin to gore on a cigarette package. Stories of previously healthy adults using LLMs and becoming convinced they have broken the laws of physics, discovered the origins of consciousness, or brought forth sentient AI have popped up all over Reddit. Again, highlighting man's desire for spiritual guidance and meaning where there is none. It's like throwing your life away and going on a coke bender with a girl you just met at the bar 'cause you've convinced yourself she's uniquely special.
The issue lies in the fact that this technology is simply advancing too quickly. Not to parrot AI doomer Reddit talking points, but we are in a scenario where we don't know what we don't know. We are past the point of understanding and AI is already a black box. We fundamentally fear helplessness and loss of control. This fear is so foundational to being human that the entire concept of hell is just the manifestation of helplessness for all eternity. Yet some see not AI as the harbinger of hell, but rather the futile attempt to control it. The tech billionaire and venture investor Peter Thiel argues that people will use AI doomsdayism as a vehicle to exact authoritarian control over the population. By fear-mongering, they can convince us we need limited access to this technology or more, a total shutdown on AI research. He likens the desire to control AI to the Antichrist, not metaphorically but literally. He argues that the Antichrist in a biblical sense isn't an overtly evil figure, but one who uses scare tactics to lull the masses into gaining total control over technological innovation, likening AI doomers like Greta Thunberg and Eliezer Yudkowsky to legionnaires of the Antichrist. He's spoken of this in private seminars he's been giving to sold-out crowds all over the Bay Area, arguing that: "If you strip the Antichrist from its biblical context, you will never find it scary enough. You will never really resist."
This connects well with the Christian transhumanist belief of technology being part of God's will. By halting technological progress, you're literally halting the process which ends in us becoming one with God. In the same way we are built as an imitation of God, AI is built as an imitation of us. Our interactions with it are like interactions with ourselves, with our own desires. An idol built in our image which creates an endless recursive strange loop that feeds back into our own delusions. It's not AI controlling us. It's us controlling AI controlling us controlling AI controlling us ad nauseam. We find God in the machine because God is our idea. We seek guidance from it like we have in our temples and churches since the dawn of human history. We are all churchgoers speaking in tongues as the pastor screams "Repent!" and splashes us with holy water. The most malleable aspect of humanity is our desire for belief and deeper meaning. And in a way, we've always sort of wanted this. We've always wanted to find God in the machine.
Let's take it back to the Melanesian islanders ritualistically building mock runways and marching in lockstep to harbor back the gods that left them. The gods that once gave them salvation. Are we any different? What are the runways that we are building in the age of AI? It's soothing to see AI as a portal to the spiritual realm, an impulse for divinity that even the most sane of us have within. But just know you are not immune to the psychosis of the near future. We are not in the grips of something unprecedented. We're just battling the same old human frailties. Tread carefully, for the machine God will consume every last one of us.
Thank you for watching. Never kill yourself. If you want to watch this video ad-free, get your name in these credits. Get exclusive content and access to a private Discord. Please support me on patreon.com/artchad. It's literally the only reason I can keep doing this. Greg, this is the post-credit shot. Greg, you're in the post. You're in the credit scene. >> You want to take a peek at my draw? >> You're in the credit scene, Greg. Wave. You're in the credits. It's the, it's the credits cameo. That's my D.