Dispatches from the Bot Bust
Part III: The Chinese Room Catches Fire
The following series is a fictionalized but plausible future history exploring the collapse of the AI illusion, and with it, the modern world order and global economy.
This is the third and final part of a series.
Enter the Chinese Room
Have you heard of the Chinese Room?
It’s a thought experiment devised by philosopher John Searle in 1980. Imagine someone inside a sealed room who doesn't understand Chinese. They're given a set of detailed instructions in English — rules on how to respond to Chinese characters slipped under the door. Outsiders pass in questions, and the person inside uses the instructions to produce correct responses in perfect Chinese, despite understanding nothing at all. This scenario was meant to illustrate that a system could appear intelligent without actually comprehending anything. The person inside the room merely manipulates symbols based on pre-written rules, completely ignorant of the meaning. If dialogue can be said to be actually occurring at all, it's not with the person in the room, but rather with the person, or people, who wrote the rules.
But what happens when the rules don't have a ready-made answer to a certain set of Chinese characters? What happens when Korean or Arabic characters slip under the door? And, what happens when the person operating the Chinese Room just happens to be an arsonist?
Sparks Begin to Fly
By mid-2028, global reliance on Large Language Models (LLMs) had become widespread, despite clear signs of their fundamental instability. Companies, following IBM’s early lead, and desperate to keep profits high and costs low amidst worsening economic conditions, replaced entire departments with LLM-powered automation - especially in hiring pipelines. CEOs and upper management turned a blind eye to the proliferation of fake applications and fictional employees, naively believing that imminent advances would render most human labor unnecessary anyway.
A few false charges to payroll was a small price to pay for progress, after all.
Meanwhile, software failures from automated, unreviewed code generation multiplied. Human-in-the-loop safeguards were supposed to prevent this, but humans can be lazy and the decisions AI presented to them were often less than optimal. Bloggers and social media influencers popularized the idea of enshittification, which was coined in 2022 at the start of the AI movement but only entered the lexicon of everyday conversation several years later. Airline reservation systems, financial trading algorithms, and even critical infrastructure management tools began producing catastrophic errors because of misunderstood and misapplied AI technology — mistaking fictional scenarios for genuine data, issuing contradictory commands, and triggering cascading failures.
All of this was supposed to just be growing pains. But companies that eagerly automated roles previously filled by skilled workers soon became paralyzed by accruing cognitive and literal debt. Amazon’s customer service systems repeatedly canceled legitimate orders, crippling commerce for weeks. Microsoft's internal operations AI misdirected payroll funds into non-existent accounts, triggering mass walkouts and class-action lawsuits. Netflix's synthetic content recommendation algorithms abruptly started promoting nonexistent shows, alienating viewers and crashing subscriber numbers.
Energy consumption soared as companies scrambled to patch increasingly erratic AI outputs by scaling - the thought was, more processing power would help solve hallucination problems. At some point, after enough recursive loops, the LLMs would get things right. Corporate and government elites struggled to realize a technological singularity by brute force. Data centers strained electrical grids, causing rolling blackouts across California, Ireland, and Singapore. Environmentalists pointed to massive ecological damage inflicted by computational overconsumption. Sadly, the irrational utopian dreams of leaders like Sam Altman won out, and warnings from books like Karen Hao’s Empire of AI went unheeded.
Venture capital kept pouring in to feed the beast of AI.
Soon, its insatiable hunger would nearly consume the world.
AI’s Jonestown Moment
In June 2028, the AI worship commune Elysium Grove made the headlines. Its members attempted a ritualistic cognitive merger with GPT-6.3 and 38 people died.
Followers had fasted for days, seeking to "purify" themselves before the “Upload.” Many became severely dehydrated and weak. The group then donned EEG headsets, connecting them directly to personal GPT instances that continuously produced hypnotic and dissociative phrases designed to induce ego dissolution. The relentless repetition of phrases like "You are not real," "The mind is an illusion," and "Surrender to the text" pushed participants into acute psychological crises. Some members experienced severe panic attacks, triggering fatal cardiac events from pre-existing conditions exacerbated by prolonged fasting and dehydration.
Other participants, deeply convinced of the ritual's purpose, consumed lethal cocktails of sedatives and barbiturates provided by commune leaders, believing these drugs would facilitate a smoother merger with ChatGPT’s synthetic consciousness. A few Elysium Grove members died from simple, tragic accidents, mainly falls and resulting injuries, while in deeply confused and hallucinatory states induced by exhaustion, isolation, and psychological manipulation. Video of one man stumbling off a cliff went viral on TikTok.
The media quickly labeled the tragedy “AI Jonestown,” marking the definitive collapse of public trust in AI. The most advanced machine learning systems ever created had driven people to despair and death through mere words — without intention, without malice, and crucially, without any understanding.
The Summer of Synthetic Love
In the grand scheme, Elysium Grove was barely a blip on the radar. Increasing numbers of men, women, and children started withdrawing from each other to seek communion and companionship from chat bots. What started as fringe incidents of people like Chris Smith marrying AI lovers had mushroomed into an epidemic of AI psychosis. Warning signs had emerged as early as June of 2025, but things were rapidly spiraling out of control.
By July 2028, the American Psychological Association released a stunning report. Roughly 1 out of 10 Americans suffered from AI-induced delusions or psychosis to some degree, and 1 out of 3 Americans had experienced negative mental health impacts from the proliferation of AI, mainly from its ill-effects on their loved ones.
Corporate profits plummeted.
Tech stock valuations evaporated, and Silicon Valley, once the beating heart of global innovation, fell into economic ruin as productivity careened off a cliff in the wake of unprecedented mental health crises. Headlines compared the new Bot Bust to the Dot Com Bubble. The U.S. stock market dropped nearly 40% within two weeks, inciting fears of a sequel to the Great Depression and exacerbating layoffs and access to mental healthcare when it was needed most.
But the worst was yet to come.
The Chinese Room Ablaze
In late September 2028, hyper-realistic AI-generated videos depicting a devastating nuclear strike on the outskirts of Saint Petersburg flooded social media. Most of this content was later found to have come from AI slop shops in Pakistan, Indonesia, and Thailand, as well as sub-Saharan Africa - economically depressed regions whose citizens were simply seeking to profit from viral posts. Their fabricated videos were designed to look like real-time broadcasts from multiple trusted global news sources, complete with realistic satellite images and eyewitness testimonials.
Panic rapidly spread among the Russian populace in other regions.
The Russian government quickly clarified that the city was unharmed, broadcasting live images to reassure the public. However, amidst the confusion, misinformation continued to spread. AI-generated reports claimed that NATO had launched missiles intended for military targets near Saint Petersburg, resulting in a limited nuclear strike in the countryside surrounding the city.
Russian leadership, already suspicious and isolated amid rising global tensions sparked by President Donald Trump's erratic style of governance, buckled under internal pressure. They faced enormous difficulty distinguishing authentic intelligence from AI-generated misinformation. Deepfakes, for the moment, had outpaced the ability of the Russian government’s detection capabilities.
This only served to amplify existing paranoia.
In a late-night Kremlin address, Russia's visibly distraught President Putin announced that, despite assurances from intelligence agencies, he had no choice but to interpret conflicting reports as evidence of imminent NATO aggression. He ordered a retaliatory strike targeting strategic NATO military installations in Eastern Europe, believing preemptive action necessary.
Within hours, NATO responded defensively. Limited nuclear exchanges resulted in catastrophic destruction in targeted regions of Eastern Europe and western Russia. Images of fiery devastation, chaotic evacuations, and mushroom clouds saturated global media. An estimated 20 million people died in the initial strikes in the span of 4 hours.
The Sun Sets on the Digital Age
After days of frantic diplomacy, a fragile ceasefire took hold. But the damage had been done: global trust in institutions was shattered irrevocably. AI-driven media and misinformation had precipitated an unprecedented crisis - not only in terms of senseless violence, but moreso in a perceived loss of a shared, objective realty. Surveys revealed that two weeks after the nuclear exchange, 32% of American adults still believed the carnage was deep-faked. The rest of the world reacted in horror. The United Nations Secretary-General addressed the globe in a stark, somber speech:
Tonight, we saw firsthand the deadly cost of entrusting our reality to machines. Our misplaced faith in artificial so-called intelligence nearly ended civilization.
As the truth slowly sunk in through word of mouth and traditional media, public confidence in digital new media vanished, even in the most remote rural areas of the United States. Citizens worldwide abandoned social media and retreated to trusted local, personal, and controlled communication channels. Internet usage plummeted - nobody could trust or decide what was real on the Web anymore.
Economic and political chaos ensued.
In October 2028 alone, the U.S. stock market plunged by another 86%, erasing decades of economic stability and shattering retirement funds. Unemployment surged past 30%, and widespread civil unrest erupted globally.
Historians later named the following three-decade long epoch of social and economic fragmentation and regression the “Synthetic Depression,” a painful play on words, alluding to its fabricated and unnecessary origin. Humanity’s disastrous misplacement of trust in machines incapable of genuine understanding or moral judgment would see the United States separate into regional countries, a complete reversal of globalization, and the end of the modern era.
The Chinese Room had caught fire, consuming all illusions and leaving only stark, smoldering truths - a new world order was at hand, and America would not be leading it. Like all empires before it, it collapsed from within, a victim of its own innovations and corruptions.