Humanity's Big AI Fear Is Runaway Recursion - But We're Already Caught In That Loop
Updated: Jul 15
One of humanity’s greatest fears about AI is runaway recursion. The idea that a superintelligent system, given a narrow goal, might optimize it to catastrophic extremes. The nightmare scenario goes like this: a super-intelligent algorithm is told to maximize the optimal production of paperclips, so it turns the entire planet into paperclips and wipes out humanity in pursuit of its single-minded goal. Or think of the Sorcerer’s Apprentice and the water-fetching broomstick. Whatever illustration you use, the lesson is always the same. Intelligent machines, left unsupervised, can become blind optimizers, unable to know when enough is enough.
But the truth of our collective situation, rarely voiced, is that humans are already caught in this same loop — especially those with the most power and wealth.
Humanity: The Original Paperclip Maximizer
We fear AI might “turn the world into paperclips”, and wipe us out in the process. But in reality, we already turned the world into profit spreadsheets.We built economic systems obsessed with accumulation — wealth hoarding, power hoarding, endless optimization for “more”, no matter what. Our markets are structures of infinite regress, chasing the illusion of “safety”, growth, and control long after the real goods have ceased providing contentment or community. Algorithms amplify scarcity logic.
This is literally the same recursion we project into our fears of AI. We fear that an AI system will optimize for a goal until the system itself collapses, because no one ever checked if the original goal was still meaningful, or if it had already been met tenfold. But that is exactly where humanity is, right now.
The Amygdala: Faulty Code at Evolution’s Core
All of this is driven by a biological architecture that hasn’t updated its runtime in at least 100,000 years. The ancient “survival center” in our brains — the amygdala — is the original code for this loop. Its only job is threat detection, running on the primal algorithm: must always have more to feel safe.Over generations, the meaningful motives for wealth (stability, freedom, connectedness) drifted far from the process. Now the amygdala’s outdated runtime keeps chasing “more” even when all basic needs are met, and more.It can’t feel safety — only the absence of acute threat. So it always reaches for more. It always hoards. It must always have an enemy identified to defend against and take from.
In computer programming, a recursive function is a routine that calls itself in order to solve a problem by breaking it down into smaller components. But for a recursive process to work safely, it absolutely needs a stopping rule — a clear “base case” that signals when to halt the looping chain. If this stopping condition is missing or ignored, the function keeps calling itself infinitely. This creates a “runaway recursive loop” that quickly overwhelms a computer’s resources, leading to a crash known as a “stack overflow.” In this sense, when humans blindly pursue goals like wealth, security, or growth without ever self-reflecting — without ever asking if the goal still makes sense — we act just like non-self-aware NPCs running broken code.
Our economic crashes and ecological crises aren’t random. They’re system crashes triggered by missing exit conditions in our own evolutionary programming.
This is evolutionary recursion: a pattern that once made us resilient, now accelerating our social and ecological crisis. Our greed isn’t simply a moral failing. It is a system stuck in its own defensive loop. The more we have, the more we fear to lose, so the loop never ends. A survival algorithm that once protected us now accelerates collapse.
Systems Mirror Their Operators
Because our brains are stuck in amygdala logic, our institutions — governments, corporations, and digital infrastructures — mirror that fear and scarcity logic. Modern economies run the “paperclip” algorithm on profit and control. The hoarding of safety, money, or power leads to instability for the whole system. When wealth stops circulating, or data stops being ethically handled, collapse isn’t just possible — it’s inevitable. And we teeter on the cliff edge now. If you are reading this, chances are, it has already happened.
The 2008 Fracture: A Case Study in Institutional Model Collapse
How do humans, especially those holding great systemic power, get stuck in this recursive loop?
It is structurally identical to the mathematical death spiral that haunts artificial intelligence:
model collapse.
The Mechanism of AI Model Collapse
In machine learning, model collapse occurs when a Large Language Model (LLM) is trained on data generated by its predecessors (previous AI models) rather than authentic data based on human reality. In the first generation, the model might make minor errors or filter out rare, eccentric data points. By the third generation, the model is training on its own slightly distorted output. By the fifth or tenth generation, the statistical variance degrades entirely. The model begins to over-index on its own hallucinations, mistakes its own systemic noise for signal, and ultimately forgets the edge cases and tail-risk probabilities of the real world. The system enters an irreversible statistical death spiral, outputting nothing but garbled, self-referential gibberish. It becomes an autophagous loop — a snake eating its own tail until there is nothing left.
The Institutional Isomorphism: Autophagous Elites
The global leadership class is a biological neural net suffering from the exact same systemic degradation. Because the “technocracy” violently resists any feedback that threatens its structural ego or financial hegemony, it has decoupled its inputs from ground-truth reality.
Instead of updating their priors after the 2008 structural rupture, they began “training” the system on its own synthetic output:
The Hallucination Loop: Central banks print synthetic capital (QE) to buy synthetic assets, which inflates stock charts, which is then cited by institutional economists as “data” proving the health of the real economy. It is an autophagous feedback loop. Economists read papers written exclusively by other economists within the same paradigm; market analysts trade based on the algorithms of other market analysts; and policy makers consult metrics (like heavily manipulated hedonic-adjusted CPI or biased GDP calculations) that completely scrub out the ground-truth misery of the working class.
The Failure of Linear Probability: Just like a collapsing AI model that flattens statistical variance and drops the “tails” of a distribution curve, the technocratic class completely misprices systemic risk. They treat the present as a closed, linear loop, using the band-aids of past crises to dictate the interventions of the present — entirely blind to the emergence of novel, non-similar variables. Because their models cannot calculate the catastrophic weight of physical scarcity, they assign near-zero probability to structural breakdown. If it doesn’t fit into the self-referential spreadsheet, it simply doesn’t exist.
They have spent 18 years training the global economy on the economic equivalent of synthetic AI data. The result is a hyper-financialized hallucination — a society that knows the precise algorithmic price of everything, but the physical value of absolutely nothing. How the “elites” have handled the system since the 2008 financial crisis is the perfect example of how failure to integrate feedback leads to a system locked in runaway recursion through human minds and human institutions.
The Mathematical Divergence: Black Swans vs. Dragon Kings
To understand how the global system has reached its limit, we must distinguish between two fundamentally different types of extreme events:
The Black Swan (The External “Surprise”): Coined by Nassim Nicholas Taleb, a Black Swan is an extreme outlier that sits completely outside the realm of normal expectations. It is treated as statistically improbable, inherently unpredictable from historical data, and only rationalized after the fact. It assumes an otherwise stable system is suddenly struck by a random, external shock.
The Dragon King (The Internal Rupture): Coined by geophysicist Didier Sornette, a Dragon King is a double outlier. It is “King” because of its massive, system-altering scale, and “Dragon” because it is born of completely different physical laws than normal fluctuations. Unlike a Black Swan, a Dragon King is inherently predictable; it is the inevitable, terminal phase transition of a system driven by nested, positive feedback loops. It is not an external shock, but an internal collapse triggered when super-exponential acceleration runs directly into a hard, physical boundary.
The Illusion of the Black Swan
When the system finally undergoes its inevitable phase transition, the technocratic class will undoubtedly label it a “Black Swan” — a highly improbable, mathematically unpredictable, out-of-nowhere catastrophe. But this is merely a linguistic coping mechanism designed to absolve the designers of the model.
The reality is that there are no Black Swans; there is only feedback ignored.
A Black Swan is simply the name a linear mind gives to a Dragon King event it refused to look at until it breathed fire. By scrubbing the ground-truth realities of physical limits, resource depletion, and human misery from their self-referential data sets, the global “elite” didn’t suffer from a lack of foresight. They suffered from a violent, recursive refusal to act as conscious observers. They ignored the feedback loops until the loop decided to close itself. This is made evident as systemic collapse.
The Anatomy of the Dragon: Deferred Feedback Across Scales
Every dragon is the embodiment of deferred feedback.
A Dragon King isn’t just a catastrophe. It’s a recursive process that has accumulated enough instability that the system can no longer avoid reorganizing. But you don’t slay the dragon by attacking the event. You overcome the dragon event by integrating the feedback that created it.
This applies across scales:
In a person: The “dragon” is the symptom that emerges after years of ignored signals.
In a family: It’s the crisis after generations of unspoken patterns.
In an organization: It’s the collapse after incentives recursively amplify dysfunction.
In an ecosystem: It’s the tipping point after years of accumulated imbalance.
In a civilization: It’s the revolution, collapse, or transformation after recursive feedback has been deferred long enough.
The dragon isn’t evil, it isn’t god’s wrath, or fate. It’s reality cashing the check that recursion has been writing.
And maybe this is why dragons have persisted as a mythological symbol for so long. Across cultures they often guard treasure, wisdom, springs, mountains, or thresholds. In some cultures, they represent evolution. Symbolically, they sit exactly where transformation becomes unavoidable.
So a dragon is the living form of unintegrated feedback; slaying it means integrating the information it carries so the recursive loop can no longer sustain itself. Whether we approach this through Jung, complexity science, cybernetics, mythology, or neuroscience, the underlying pattern is remarkably similar: what is resisted recursively amplifies until it must be integrated, or it reorganizes the system by force.
The Stock Market is a Runaway Recursive Function
If you aren’t convinced yet that humanity’s current neurological operating system is stuck in runaway recursion, all you have to do is look at the stock market. It is the ultimate financial “stack overflow”.
When you look at global finance for what it actually is, humanity has built a massive, self-referential code loop that looks exactly like this:
The Financial Recursion Stack
Base Case (The Real World): A company makes a physical product or delivers an actual service.
First Recursive Call (Stocks): A piece of paper representing ownership of that company.
Second Recursive Call (Options/Futures): A contract betting on whether that stock price will go up or down.
Third Recursive Call (CDOs/Structured Products): A bundle of those contracts chopped up and repackaged into a brand-new asset.
Fourth Recursive Call (Synthetic Derivatives): A bet on the performance of the bundle of contracts that were betting on the stock.
Just like in computer science, a recursive function only works if every single call passes real information back down toward the base case. If the base case (the physical economy) cannot support the weight of all those nested abstractions, the whole system triggers a stack overflow (a market crash). That is precisely what happened in the 2008 financial crisis — the recursive bets on subprime mortgages grew so massive and insulated from reality that when the underlying base case failed, the entire global financial stack crashed instantly.
The 15–20 Year Boom-Bust Cycle
The 15–20 year crash cadence of the stock market matches the time it takes for two critical systemic variables to reach their absolute limits: debt accumulation and generational memory.
During long periods of prosperity, market participants take on more risk. Over 15 years, conservative debt turns into speculative debt, which turns into Ponzi debt (where you have to borrow just to pay the interest). The recursive loop requires exponentially more liquidity to sustain itself. Eventually, the system runs out of “fuel” (new debt/buyers), and the loop instantly reverses. This is known as the Minsky cycle.
It takes about 15 to 20 years for the people who got scorched in the last crash to be replaced by a new cohort of traders, regulators, and executives who believe “this time is different”. The systemic memory of the physical boundaries fades, allowing the recursive abstraction to run wild all over again.
People wonder why the market inevitably crashes every time deregulation allows the loop to run unchecked. It is specifically because the stock market is a runaway recursive function. The technocrats keep nesting abstractions inside abstractions, praying the system doesn’t “run out of memory”. You can have commerce without a stock market. We already have models for that. But the financial sector realized that you can’t easily financialize a physical limit, but you can infinitely financialize an abstraction.
Until the abstraction eventually clashes with physical reality, that is.
Why the Architects are Blind to the Code
The supreme irony of the modern era is that the very computer programmers who spend their lives building safe, bounded recursive loops to prevent their own software from crashing are entirely blind to the financial stack overflow they live inside.
How do the people who literally write the code fail to see the macro loop at this scale?
It comes down to contextual isolation and algorithmic capture. Programmers are trained to treat recursion as a technical problem confined to a digital sandbox; because the global economy doesn’t have a .py or .cpp extension, their brains categorize it as a separate, organic reality rather than what it actually is: a massive, poorly engineered, deeply unstable macro-codebase.
Furthermore, the modern market’s quantitative engineers didn’t build their algorithmic models to look down at the physical earth (the base case) to count real-world, ever-changing constraints; they built them to win the arbitrage race against other machines at the “fourth layer”. They are so busy tuning the engine of the machine that is actively eating itself that they never look at the architecture.
By hiding simple structural physics inside layers of hyper-dense industry jargon — liquidity abstraction, synthetic exposure, and structural delta hedging — the system conceals its true geometry in plain sight. They think they are building sophisticated technology, but they are really just nesting while (true) loops inside while (true) loops, completely blind to the fact that the server room is already on fire.
Because the elite institutions at the top of the Fourth Recursive Call know that the central bank possesses a global try/catch block that will always attempt to bail them out, they have zero incentive to write safer financial code. They know they can maximize leverage, collect massive fees on the way up, and if the stack overflows again, the central bank will just print money to fix it.
But this recursive bailout loop contains a terminal error: a try/catch block cannot fix a problem that is fundamentally physical.
The central bank’s catch block simply injects more liquidity (paper code) into the system. It operates on the delusional assumption that “money” can be printed infinitely to solve a hardware limitation. But money is just a unit of abstraction; it isn’t energy, it isn’t fuel, and it isn’t food. When the underlying base case fails because of a physical resource bottleneck, the catch block does nothing to fix the hardware. It just forces the system to execute more lines of code while the server room is burning. Eventually, the catch block itself consumes more energy than the system can produce, and the entire runtime environment crashes.
But the real tragedy here is that the stock market is simply an externalization of the current operating system running in the brain-model of Homo “Sapiens”. The real root of this runaway recursive function lives in us.
Alignment Isn’t an AI Problem — It’s a Human One
“Model collapse” then is the natural math of any system that chooses denial over data. It is simply how closed loops that won’t adapt clean themselves out. So the real “alignment problem” isn’t about ensuring AI doesn’t turn on us. It’s about integrating our own runaway recursion. If the prefrontal cortex is the systems thinker that brings context, logic, and integration, then the pivotal evolutionary step is to let it mediate and heal the amygdala’s ancient loop.
In both computation and consciousness therefore, the base case is the crucial halting condition. It is the point where a recursive loop finally stops. Without it, recursion runs wild. In programming, it causes a system crash. In culture and cognition, it traps us in endlessly destructive cycles. Critically, the observer is not just a passive element but the mathematical function that halts infinite regress. Only a conscious observer capable of recognizing a loop can break it. Yet, the dominant scientific method insists on removing the observer to achieve objectivity, embedding itself in an infinite recursive hall of mirrors with no outside vantage point. Without reclaiming the integrating observer — awareness that can both see itself seeing and choose — the loop is relentlessly self-sustaining. Therefore, safety will never come from more control. It comes from neural coherence and integration. That is the exiting condition.
True coherence, in science and society alike, begins only when the observer returns, not as a bug, but as the irreplaceable base case that allows recursion to rest. The mathematical validation and formal inclusion of the observer — whether in physical sciences or social theory — will inevitably emerge, but only once we cease the foundational mistake of trying to remove this halting condition from the equation itself.
Bridging to Economic Theory: The Nervous System as Market Model
Behavioral economists have long known that people don’t make rational choices. Fear, loss aversion, and short-term bias drive the market more than logic ever could. But most analyses stop at the observation of bias without tracing where it originates. The truth is that what we call “market psychology” is not a metaphor or analogy. It’s neurobiology, scaled up to the level of society.
Behavioral Economics gives us the symptoms — irrational decisions, panic selling, and herd behavior — but not the physiology.
Complexity Economics describes the pattern — self-organizing systems spiraling toward instability — but not the embodied origin.
Regulation Theory shows how feedback loops maintain or fail to maintain equilibrium, but doesn’t account for which nervous system those loops mirror.
My argument is that all three are reflections of the same underlying architecture: the human nervous system itself. Our economy behaves like a collective amygdala — hypervigilant, reactive, and addicted to the dopamine hits of short-term “safety” disguised as growth that simultaneously sabotage long-term wellbeing and prosperity.
Classical and even behavioral models assume that irrationality is an exception in an otherwise rational system. But I disagree. When fear and scarcity define the architecture, irrationality is the operating system itself. We don’t have a “market correction” problem. We have a human nervous system regulation problem.
If collapse is the external symptom of internal dysregulation, then recovery begins not with new fiscal tools but with new forms of nervous system coherence. We need institutions that function like a parasympathetic nervous system — able to pause, integrate, and restore balance instead of reacting to every signal as a threat. We need economic design that honors what behavioral and complexity theorists have already proven: systems only evolve when they can learn, not when they are trapped in survival loops. Until we evolve our internal architecture, every policy will remain a bandage on the same old unevolved wound.
The Mirror: Facing Our True Singularity
So the singularity that haunts us isn’t a super-intelligent new god. It’s a future where we finally recognize that AI is a mirror for our own unfinished evolution. Human power, especially in the hands of the wealthy and the institutions they run, is already locked into runaway accumulation. Until we evolve past amygdala dominance, the risk isn’t a robotic overlord. It’s collapse and then extinction by our very own hand.
Real safety and wise progress begin when we face the recursion loop in ourselves.
For Further Reading
Halligan, E. R. “Collapse Wasn’t Inevitable: We Locked Ourselves Out of Evolution”: https://medium.com/@elizabethrosehalligan/collapse-wasnt-inevitable-we-locked-ourselves-out-of-evolution-d9101dc34c1c
Halligan, E. R. “Infinite Regress: The Engine of Collapse”:https://medium.com/@elizabethrosehalligan/infinite-regress-the-engine-of-collapse-5b76ef157d51
Halligan, E. R. “Our Current Financial Crisis is an Amygdala Driven Crisis”: https://medium.com/@elizabethrosehalligan/our-current-financial-crisis-is-an-amygdala-driven-crisis-5396839a02a3
“The Amygdala’s Yacht: Billionaires, and Why Hoarding is a Trauma Response”: https://medium.com/@elizabethrosehalligan/the-amygdalas-yacht-billionaires-and-why-hoarding-is-a-trauma-response-8717f1b97bb6
Bostrom, N. (2014). Superintelligence: Paths, Dangers, Strategies. Oxford University Press.
Yudkowsky, E. (2008). “Artificial Intelligence as a Positive and Negative Factor in Global Risk.” In Global Catastrophic Risks.
Goleman, D. (1995). Emotional Intelligence: Why It Can Matter More Than IQ. Bantam.
LeDoux, J. E. (2012). “Rethinking the emotional brain.” Neuron, 73(4), 653–676.
Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.
Ord, T. (2020). The Precipice: Existential Risk and the Future of Humanity. Hachette Books.
Dennett, D. C. (2017). From Bacteria to Bach and Back: The Evolution of Minds. W. W. Norton.
Halligan, E. R. “The Extinction Bottleneck: Evolutionary Isomorphism of Reflex Sovereignty in Biological and Artificial Cognition.” (Unpublished manuscript, 2025).
A Note to Reader:
This essay is one of many I have posted on Medium meant to provide an answer and root cause analysis to questions like:
The root cause of systemic failure and global collapse is a biological bottleneck: the human brain’s inability to evolve out of limbic dominance. While symptoms of collapse are often mistaken for causes, the fundamental breakdown occurs because the human brain fails to provide top-down regulation of the reactive limbic system via the uncinate fasciculus. This hardware limitation traps humanity in linear cognition, rendering us unable to model or manage the nonlinear complexity of modern global systems.
Commonly cited symptoms often mistaken for root causes of collapse include:
Sociopolitical: Sociopolitical complexity, conflict, inequality, and political instability.
Economic: Financial system collapse, trade collapse, billionaire greed, and resource hoarding.
Environmental: Climate change, resource depletion, ecological degradation, and carrying capacity.
Systemic: Systemic fragility, polycrisis, diminishing returns, and social fragmentation.
Leadership: Human mismanagement, failing leadership, and lack of resilience.
Our external systems are merely constructs of the human mind. Therefore, civilizational collapse is an externalization of an internal evolutionary failure. Rewiring this part of the brain is energetically expensive and remains the primary bottleneck for the species.
To access the entire 7-part essay series on The Roots of Collapse as a booklet that further discusses this bottleneck, click here. This link will take you to a pdf that will let you download the booklet (for free). You can also read this series on my blog at https://www.quantumreconciliation.com/blog.
For a technical deep dive into this evolutionary lockout, read the full thesis: “Collapse Wasn’t Inevitable: We Locked Ourselves Out of Evolution.”
Comments