The Scientific Taxonomy of Mind: Intelligence vs. Stupidity
Thur, July 09 2026 /Mpelembe Media/ — According to complex systems theorist David C. Krakauer, the common assumption that stupidity is merely a passive vacancy of thought is fundamentally incorrect. Standard psychometrics, such as IQ tests, attempt to reduce cognitive value to a single price-like metric, which is as reductionist as trying to understand the artistic value of a Picasso masterpiece solely by its auction price.
Instead, a robust taxonomy of cognitive states requires separating four distinct regimes:
- Ignorance: A simple insufficiency of data. Like a random-walk algorithm, it is highly inefficient but will eventually stumble upon a correct answer over a long time horizon as data fills the void.
- Error (Being Wrong): The misapplication of a rule due to sensory noise or temporary distortion. Because the system maintains epistemic flexibility, it remains data-responsive and can quickly correct its trajectory.
- Intelligence: The deployment of elegant, highly adaptive rules that compress the problem space, solving complex problems with minimal effort and a positive marginal return on data.
- Stupidity: The active, rigid application of a flawed rule system. Stupidity is defined as using a rule where adding more data does not improve the chances of success, and indeed, makes outcomes progressively worse.
Because stupidity requires an impressive apparatus of complicated theories, elaborate beliefs, and sophisticated algorithms to turn a simple problem into an exceptionally difficult one, stupidity is a highly sophisticated behavior that scales directly with intelligence. A rock, a river, or a thermostat cannot be stupid because they lack the representational capacity to build and defend an incorrect model of reality. This pathology is mirrored in Kahneman’s dual-process theory: stupidity manifests when fast-thinking System 1 heuristics lock onto a flawed rule, and analytical System 2—instead of acting as an objective error-correcting mechanism—is co-opted to construct sophisticated rationalizations to defend the error.
The Unconscious Mind and the Parasitic Nature of Language
A parallel exploration of human cognitive evolution is detailed in novelist Cormac McCarthy’s non-fiction essay, “The Kekulé Problem”. McCarthy defines the unconscious as a highly competent, ancient biological “machine for operating an animal” that has successfully directed life without language for millions of years.
The central paradox—the “Kekulé Problem”—asks why the unconscious communicates through dreams, pictures, and parables (such as chemist Friedrich August Kekulé dreaming of a snake eating its tail to solve the ring structure of the benzene molecule) instead of using direct verbal language. McCarthy and Krakauer propose that language behaves structurally like an infectious parasitic invasion. Because language is an evolutionary newcomer that bypassed traditional developmental pathways to colonize the least dedicated regions of the cerebral cortex, the ancient animal unconscious does not trust or like it. While language is an exceptionally useful tool for collective computation, it also acts as a self-propagating vector that can trap human populations in shared, highly coordinated ideological distortions.
The Algorithmic Frontier: Artificial Intelligence and “Super-Stupidity”
As humanity integrates large language models (LLMs) and artificial intelligence, the threat of collective cognitive collapse rises. AI is engineered to minimize human cognitive effort, but this creates a dangerous loop: the more humans rely on frictionless tools, the less they engage their own deliberate System 2 reasoning, causing natural cognitive capacity to degrade.
This algorithmic landscape introduces the era of super-stupidity. AI hallucinations—where models produce fluent, detailed, and highly confident text containing completely fabricated references and facts—are not easily patched bugs, but fundamental structural features of transformer architectures that optimize for the plausibility and appeal of their outputs rather than objective truth. It is the ultimate realization of what Robert Musil called “intelligent stupidity”—a flamboyant, systematic generation of fluent nonsense executed with immense computational power. The ultimate hazard is that as modern systems exceed human cognitive limits, the vital task of error-detection is outsourced to machines, leaving human decision-makers incapable of identifying highly complex, mathematically opaque computational errors.
Strategic Friction and Pluralistic Governance
To mitigate these systemic threats, complex systems theory suggests implementing “cognitive friction”. To keep analytical reasoning active, cognitive tools must be designed to enhance rather than replace human thought by introducing deliberate, strategic obstacles. In his personal workflow, Krakauer exemplifies this by rejecting frictionless convenience, instead using split mechanical keyboards (ErgoDox EZ and X-Bows) as “unforgiving taskmasters,” alongside the steep learning curves of Emacs, Mathematica, and LaTeX to slow down his cognitive process.
Furthermore, to combat institutional collapse, global governance must shift away from demanding shared, uniform ideological values—which only worsens polarization. Instead, institutions should model themselves on pluralistic biological ecosystems, where thousands of highly diverse species with different objectives exist in relative harmony by maintaining specialized niches and localized cooperative interactions.
Why Stupidity is the “Evil Twin” of Intelligence (and Why AI Might Make it Worse)
The Hook: Why Being Smart Isn’t Enough
We inhabit a century of staggering cognitive paradoxes. Human beings possess the collective brilliance to land autonomous rovers on Mars, sequence the human genome in a matter of hours, and engineer circuits at a nanometer scale. Yet, despite these triumphs of logic, we remain profoundly susceptible to conspiracy theories, institutional collapse, and catastrophic systemic failures. We are a species that can solve the secrets of subatomic particles but cannot manage a stable pandemic response or an equitable economic system.Why do “smart” people and highly sophisticated civilizations make such disastrous errors? According to David Krakauer, President of the Santa Fe Institute and a featured speaker at the Cornell Systems Thinking Conference, the answer lies in a fundamental misunderstanding of what stupidity actually is. In Krakauer’s view, stupidity is not a vacancy of thought or a low-IQ state. It is a sophisticated, high-order behavior—a “law of nature” that scales directly with the complexity of the system. To understand our current crisis, we must view stupidity as a “Science,” a necessary counterpoint to our study of intelligence.
Takeaway 1: Stupidity is a High-Order Skill
To understand the danger we face, we must first recognize that stupidity requires “sophisticated machinery.” From a systems perspective, a rock, a river, or a bacterium cannot be stupid. These entities simply obey the laws of physics or basic biological heuristics without the capacity for representational thought. As Krakauer notes, stupidity implies a capacity for getting things right before it can get them spectacularly wrong.Stupidity occurs when the slow, analytical “System 2” reasoning—our capacity for deliberate thought—is co-opted to act as a high-priced lawyer for the flawed heuristics of “System 1.” Instead of correcting an error, the intelligent mind uses its resources to construct elaborate, institutionalized systems of falsehood. We see this in the 1996 Everest disaster, where experienced climbers used their considerable expertise to rationalize staying on the mountain long after their own rigid turnaround rules had been violated. This “intelligent stupidity” is far more dangerous than simple ignorance; it is the flamboyant marshalling of the intellect in the service of a catastrophe.Stupidity is not the opposite of intelligence but its evil twin, the dissimulating Cain to a cerebral Abel. And perhaps surprisingly, the degree of stupidity available to any system scales directly with the intelligence that system possesses—more intelligence begets greater feats of stupidity.
Takeaway 2: The Rubik’s Cube Taxonomy
To move toward a rigorous “science of stupidity,” Krakauer proposes a clear taxonomy of cognitive states using the analogy of a Rubik’s Cube. With its 43 quintillion configurations, the cube represents the “rugged fitness landscape” of a complex problem:
- Ignorance: Attempting to solve the cube through random, unguided twists. While it might take several lifetimes, a random walk on a connected graph will eventually reach the solution.
- Error: Making a mistake due to a misapplied rule or sensory distortion. Because an erroneous system remains epistemicly flexible, the introduction of new data allows the user to recalibrate and return to the correct path.
- Intelligence: Deploying an elegant set of rules or algorithms that solve the cube in n steps or less, making a hard problem appear effortless.
- Stupidity: Systematically rotating a single face of the cube forever. Because the agent is following a rigid, flawed rule, they are trapped in a cyclic loop. Adding more data only reinforces the failure.We see this dynamic in nature with the moth’s transverse orientation . For millions of years, maintaining a fixed angle to a celestial light source was an “intelligent” rule for straight-line navigation. However, when humans introduced artificial lights, that same rule became a “stupid” suicidal loop, forcing the moth to spiral into the flame.
Takeaway 3: The “Parasitic” Nature of Language
The evolution of human intelligence is inseparable from the evolution of language, a concept explored in the “Kekulé Problem”—a collaboration between Krakauer and the late novelist Cormac McCarthy. McCarthy observes that while the unconscious is a “machine for operating an animal” that has functioned for millions of years through images and parables, language is a recent, “parasitic” upstart.McCarthy references the chemist August Kekulé, who solved the configuration of the benzene molecule after dreaming of a snake eating its own tail—the ouroboros . The question is: why did the unconscious use a mythic image rather than just telling him it was a ring? The answer lies in the antiquity of the “Night Shift” (as physicist George Zweig calls it). The unconscious has been getting along quite well without language for two million years; it does not trust the upstart.Language allowed for “exbodiment”—the ability to store our cognition in the world (books, culture) rather than just our brains. However, this also created a vulnerability where we can become trapped in “kayfabe,” a performative structure where we maintain shared illusions through language, similar to the Academy of Lagado in Gulliver’s Travels , where scholars spend decades trying to extract sunbeams from cucumbers.The unconscious is a machine for operating an animal.
Takeaway 4: Beware of “Super-Stupidity” in AI
The rise of Artificial Intelligence introduces the threat of “super-stupidity.” Large Language Models (LLMs) are not truth-engines; they are probability engines optimized for plausibility . When an AI “hallucinates,” it is demonstrating the pure form of Robert Musil’s “intelligent stupidity”—the flamboyant, systematic generation of fluent nonsense .The danger lies in a “cybernetic trap,” similar to Project Cybersyn in 1970s Chile. Cybersyn attempted to manage a national economy—a non-linear, complex adaptive system—using a centralized feedback model designed for simple machines. It overcomplicated the problem by trying to model millions of adaptive agents through a rigid room of telex machines.When we transition from Cognitive Enhancing Tools to Cognitive Replacing Tools , we risk a civilizational collapse:
- Enhancing Tools (like a telescope) require active learning and expand our capacity.
- Replacing Tools (like AI text generators) minimize effort and induce a dependency.If the machine’s errors exceed the user’s capacity to detect them, we lose the ability to catch mistakes that a child—or a more “ignorant” system—could easily see.
Takeaway 5: The Cure is “Cognitive Friction”
To resist the slide into “intelligent stupidity,” we must intentionally reintroduce “cognitive friction.” Krakauer argues that “efficiency is a dirty word” when it leads to cognitive atrophy. Our tools should behave like “unforgiving taskmasters.”In his own practice, Krakauer utilizes tools that force the brain into “System 2” deliberate reasoning. He conducts his work using:
- Split mechanical keyboards (like the ErgoDox EZ) that act as physical puzzles.
- LaTeX and Emacs , which require high-friction, structural interaction.
- Analog synthesizers , which provide physical limits compared to the frictionless ease of digital automation.By making the task mentally demanding, we ensure that our error-correction mechanisms remain engaged. We must build “cognitive speed bumps” to prevent our minds from surrendering to the immediate ease of automated, and potentially stupid, solutions.
Conclusion: Building a Pluralistic Ecosystem
The Cornell Conference theme, “Any Person, Any System,” reminds us that we must view humanity as a complex ecosystem. In biology, diverse species with conflicting objectives live in relative harmony through niche structures. We should not aim for a monoculture of shared values, which often deepens polarization, but for institutions that support diversity.As we navigate an era of “intelligent stupidity,” our most vital task is to distinguish between the forgeries of the mind and authentic insight. Like the protagonist in William Gaddis’s The Recognitions , who expends more effort on forged Flemish paintings than an original work would require, we often invest our greatest intelligence into our most fraudulent constructions.As you navigate your digital life today, ask yourself: Which of your “convenient” tools is actually acting as a parasitic agent, removing the friction necessary for you to see the flame before you spiral into it?