Copyright Forbes
How a reactionary idea about “efficiency" slipped into AI’s operating logic, and why leaders must challenge it. Every era mistakes its tools for its truths. The printing press made knowledge seem permanent; the internet made it infinite. Artificial intelligence has made it efficient, and in doing so, it has ushered in a Dark Enlightenment, an age when optimization has replaced wisdom as the highest good. What began as a search for intelligence has become a cult of efficiency. The phrase Dark Enlightenment comes from a reactionary 2010s current that rejected equality and democracy in favor of hierarchy and control, a warning about how easily “efficiency” can become an ideology rather than a tool. You can see it everywhere, from the halls of government to the boardrooms of companies rushing to prove they’re “AI-first.” In July 2025, the White House released its AI Action Plan, promising that artificial intelligence would boost U.S. productivity and “reinvigorate innovation.” Markets cheered as firms echoed the line, announcing “AI-driven efficiency” initiatives and workforce “optimization.” Labor economists quickly noted that measured job effects were still modest and uneven. The numbers looked confident, the reality felt ambiguous. AI has not yet made companies much smarter; it has simply made their stories more efficient. That is the essence of the Dark Enlightenment, a reactionary philosophy born online that has quietly become the operating logic of the modern economy. Artificial intelligence is not just accelerating decisions; it is redefining what counts as knowledge, progress, and even reality. Reality Drift: Inside the Dark Enlightenment of AI Across hundreds of digital ecosystems, AI models now learn primarily from the reflections of their own predictions. Each cycle of training feeds on prior outputs, compressing the world into ever-tighter statistical loops. Reality drift is the condition that follows: the moment when a system’s internal coherence outweighs its external accuracy. It no longer learns about reality; it learns instead of reality. When that happens, feedback becomes the product. Patterns feed patterns, meaning collapses into repetition. At scale, truth turns from a collective agreement into a computational coincidence. In Denmark, Jakob Engel-Schmidt, the country’s Minister for Culture, has become one of the first government officials in the world to propose legislation giving citizens legal ownership of their own likeness, their face, voice, and bodily features, in response to AI-generated deepfakes. Engel-Schmidt argues that the bill is about more than image rights; it is about re-establishing reality itself as a public good. “In the bill we agree and are sending an unequivocal message that everybody has the right to their own body, their own voice and their own facial features, which is apparently not how the current law is protecting people against generative AI,” he said. Coming from a nation known for digital governance and privacy leadership, Denmark’s move matters because it treats identity as intellectual property, effectively creating a new category of personal copyright. It signals a world where governments are being forced to legislate not only against misinformation but for ontological integrity: the right to remain real. The effort to restore what is real is not confined to lawmakers. Industry leaders are confronting the same question: how to keep progress from hollowing out purpose. Across the stack, you can watch performance turn into a proxy for truth, and that is exactly what the data shows. GUDEA, a narrative intelligence platform and data research firm that maps how information behaves across digital networks, tracks the feedback loops between human behavior and algorithmic response. Their researchers study what happens when predictive systems start learning from their own outputs instead of the world they were built to model—a shift with enormous implications for brands and institutions built on trust. Keith Presley, Founder and CEO of GUDEA, told me “When AI becomes the arbiter of truth in a fragmented information environment, we risk mistaking performance for objectivity. The underlying data can be manipulated, layered with coordinated, inauthentic content, creating a reality that is optimized but not true. Leaders need to understand that without guardrails, large models will drift toward the loudest signals, not the most accurate ones. We’re early enough in this evolution to design protections—systems that prioritize verified content, detect narrative manipulation, and keep truth in the loop before public trust is permanently outsourced to machines. “This moment demands humility and precision. AI doesn’t ‘know’ what’s true; it performs based on what it’s fed. And right now, the internet is not a neutral source of truth; it’s full of coordinated noise and manipulated information. Business leaders should assume their brand’s identity, online and inside AI systems, is constantly being reshaped, for better or worse. The hopeful part? We can intervene. With the right tools, we can detect manipulation, correct drift, and design AI systems that amplify trust rather than erode it.” In this new environment, brand health becomes a function of ontological health, how well a company can tell the difference between human and non-human activity online, and how quickly it can correct for synthetic distortion before it becomes truth by repetition. This drift is already visible in the way AI amplifies persuasion and influence. As Forbes has reported, AI-driven dark patterns are redefining manipulation itself, optimizing not for accuracy or empathy, but for engagement at any cost. The same phenomenon is appearing in business and governance. The dashboards still light up green, yet the lived world feels off-axis. From Reason to Recursion: The Philosophical Roots of the Dark Enlightenment The term “Dark Enlightenment” entered circulation in 2012, coined by British philosopher Nick Land and developed further by American technologist Curtis Yarvin, better known by his online alias Mencius Moldbug. As Forbes noted in its coverage of accelerationist ideology influencing the tech elite, Land’s ideas have quietly shaped parts of Silicon Valley’s worldview, where speed and disruption are treated as moral goods Their ideas are a warning: they show how easily the language of optimization can slide into the politics of exclusion. They argued that democracy and equality were outdated operating systems and that governance should be recoded for efficiency and control. Yarvin imagined a “CEO-state,” Land urged “intelligence pressed against the limits of control.” Together they sketched a blueprint for technocratic control, dressing hierarchy in the language of progress, replacing deliberation with data, politics with optimization, humans with systems. Although no American politician campaigns under the banner of the Dark Enlightenment, its influence has quietly threaded through the tech-populist right. Reporting has linked Curtis Yarvin’s writings and Nick Land’s accelerationist ideas to figures such as J.D. Vance and networks around Peter Thiel, where calls to “run government like a start-up” echo the movement’s vision of efficiency as virtue. Their ideas have influenced parts of the alt-right, but the deeper contagion was conceptual. The logic of the Dark Enlightenment, speed, scale, hierarchy, escaped ideology and embedded itself in management theory. From Washington’s “lead-at-all-costs” competitiveness rhetoric to progressive calls for AI regulation framed around fairness and inclusion, both sides of the political aisle now share a single technocratic instinct: engineer outcomes first, retrofit legitimacy later. You can see it across the spectrum. The right’s reflex is control without empathy; the left’s is empathy without accountability. Each, in its own way, believes the system can be engineered into virtue. Both rely on optimization as their moral grammar. The Industrialization of Intelligence: Scaling the Dark Enlightenment We are entering a cycle of mass consolidation in consulting, marketing, and technology services, not driven by invention but by automation optics. Companies are performing productivity. AI is deployed to streamline headcounts and signal discipline to investors, even when measurable ROI remains uncertain. In effect, we are witnessing human capital as the new liquidity event: firms signal value by collapsing headcount and declaring “AI-driven productivity gains,” even as researchers debate whether those cuts create real innovation or only performance optics. Beneath that spectacle lies the industrialization of intelligence. Anything procedural is being absorbed by models; what remains, intuition, creativity, judgment, is scattered across smaller, adaptive ecosystems of independent operators and hybrid human-AI teams. Executives call this optimization. In truth, it is cognitive off-shoring, a hollowing-out of the creative middle class. Economic drift is the financial twin of reality drift. As organizations chase algorithmic optics, financial models begin to learn from their own projections. Markets reward efficiency narratives detached from real output. Both drifts optimize for appearance, not substance, and both mistake repetition for truth. And repetition is where the human mind becomes the machine’s unwitting accomplice. AI and modern marketing both exploit what psychologists call the availability heuristic, our tendency to mistake frequency for validity and vividness for truth. When people encounter a message often enough, they start believing it not because it is right, but because it is remembered. Brands already understand this intuitively. Flood the zone with content, and the mind confuses ubiquity with authority. When AI joins the loop, the effect multiplies: synthetic repetition turns probability into perceived fact. The more something is seen, the more it feels true, and in the absence of friction or reflection, familiarity becomes proof. This is the new epistemic vulnerability. In an environment where attention is optimized and cognition is outsourced, availability becomes the new credibility. The result is a market, and a society, where the loudest signal outcompetes the most accurate one. A Brief History of Hollowing: How the Dark Enlightenment Hollowed Meaning The nineteenth-century industrial revolution automated muscle. The twentieth automated memory. This one automates meaning. Each wave began in invention and ended in consolidation: railroads, steel, telecom, dot-com. Each time, giants claimed inevitability while smaller firms quietly built the next world in the shadows. The pattern is always the same: innovation opens the frontier, consolidation fences it. Industrialization once hollowed out craft; digitization hollowed out expertise. Now, algorithmic automation is hollowing out authorship itself. When intelligence becomes a service, every discipline risks becoming infrastructure. Creativity, journalism, medicine, finance, each trades the nuance of judgment for the convenience of scale. AI will repeat the pattern, only faster, because what it automates now is not motion or storage; it is thought. And when thought is industrialized, the scarcity shifts. The rarest resource is no longer energy or information; it is interpretation. The capacity to connect facts to meaning, and meaning to purpose, becomes the last form of competitive advantage left to humans. The danger is not that the machines will take our jobs. It is that they will take our confidence, the belief that understanding still matters. The Efficiency Mirage: The Economic Engine of the Dark Enlightenment Automation has become performance art. AI today is more often used to signal innovation than to create it. The spreadsheet looks cleaner; the narrative pleases analysts; the intelligence layer, the messy interpretive work that links data to understanding, shrinks. When optimization becomes ontology, metrics replace missions. The means become the meaning. A civilization optimized for performance is one that has quietly given up on understanding. Hidden inside that logic is an older economic trap: Jevons Paradox. In 1865, the economist William Stanley Jevons noticed that as steam engines became more efficient, coal consumption did not fall; it exploded. Cheaper energy expanded its own demand. Efficiency produced acceleration, not restraint. AI is replaying that paradox at planetary scale. Every model that automates cognition does not replace human effort; it multiplies the number of problems deemed solvable. We do not save attention; we consume more of it. We do not reduce labor; we abstract it, replicate it, and reclassify it as data. The promise of efficiency becomes its own accelerant, expanding the system it was meant to streamline. This is the Jevons Loop of Intelligence, where each gain in computational efficiency triggers exponential demand for computation itself, deepening dependence on the very machinery that was meant to liberate us. The result is not emancipation but meta-efficiency, optimization for its own sake, where every improvement expands the frontier of exhaustion. It is the economic engine of reality drift: the more efficiently we process information, the less time we spend questioning whether the information is true. The paradox explains why AI’s efficiency boom has not delivered its promised dividend. Like Jevons’s coal economy, today’s intelligence economy consumes what it refines. Optimization feeds appetite. The cost of thought falls toward zero, so we think in wasteful abundance while meaning, the rarest resource, becomes scarce. That scarcity is the paradox’s ultimate dividend: abundance that impoverishes understanding. From the far left, critics argue this is not drift at all but design. As Shoshana Zuboff wrote in The Age of Surveillance Capitalism, the data economy did not democratize knowledge; it privatized it, enclosing human experience the way factories once enclosed labor. Cory Doctorow calls the outcome “enshittification,” the stage where platforms degrade the very systems they mediate once profit replaces purpose. And Sarah Chander, of Equinox: Racial Justice Initiative, warns that AI has become “a technology of inequality,” industrializing bias while concentrating power in fewer hands. Taken together, these voices frame AI not as a neutral innovation but as the latest phase of capitalism’s automation logic, the extraction of meaning itself. In this view, the Dark Enlightenment is not a philosophical glitch; it is capitalism’s most efficient upgrade, a system that measures everything except what matters. Two Faces of the Same Machine: Political Expressions of the Dark Enlightenment The authoritarian right and the paternalistic left appear to be opposites but share a chassis: algorithmic rule. Both see in technology the chance to perfect what politics never could, order without debate, virtue without contradiction. The right seeks order through control, the CEO-state. It imagines society as a balance sheet, a hierarchy of efficiency where dissent is noise and the market is the measure of truth. Its dream is seamless governance, frictionless obedience, a dashboard democracy, where everything is optimized and nothing is questioned. The left seeks virtue through design, the moralized algorithm. It imagines a system so fair and ethical that bias can be engineered out, harm modeled away, and justice coded into being. Its dream is benevolent automation, a world made safe by calibration. Each believes its machine can correct the flaws of human nature. Each forgets that error, disagreement, and ambiguity are not bugs in the human program but the source code of democracy itself. In both stories, optimization takes the throne and citizenship is recast as compliance. Both compress pluralism into code and erode the friction, doubt, dissent, and delay that keep society human. If the right’s Dark Enlightenment manifests as reality drift through authoritarian clarity, a world run by prediction and profit, the left’s mirror creates reality drift through moral certainty, a world run by safety and virtue. One demands obedience; the other demands consensus. Both produce compliance. The difference is aesthetic, not structural. One centralizes control in the name of strength; the other centralizes virtue in the name of safety. Either way, decision-making migrates from conscience to computation. Whether through markets or morality, both sides confuse management with meaning. When ideology becomes code, governance turns into automation. After the Purge: Escaping the Dark Enlightenment Every drift generates its counter-current. When everything repeatable is automated, meaning becomes scarce. The rebound will come from those who preserved friction: independent thinkers, small labs, hybrid human–AI guilds that kept interpretation alive. When the efficiency bubble finally bursts, what people will crave is not speed; it is sincerity. They will rediscover what machines cannot deliver: trust, intuition, and moral imagination. The next creative and strategic decade will belong to this recovery zone, a network of artisans, analysts, and builders who treat intelligence as collaboration, not substitution. We will see the rise of interpretive enterprises: studios, consultancies, and collectives built around judgment, context, and meaning-making. In a market saturated with instant answers, the scarce commodity will be understanding, patient, explainable, and accountable. Capital will chase this shift. Just as the industrial era rewarded those who mastered production, the post-automation era will reward those who master discernment. The differentiator will not be how much data you can process but how much ambiguity you can tolerate without losing the thread of truth. The new innovators will not compete with AI; they will choreograph it. They will use the machine’s pattern recognition to surface the unknown, not to replace the human in the loop but to deepen the loop itself. Friction will become a premium, the new authenticity marker in a world flooded by frictionless artifice. The story is not AI replaces; it is AI resets the hierarchy. And the next hierarchy will be built by those who understand not only how to move fast but how to think long. At the top of that new order will stand the rarest capability left: the ability to mean what you make. What Leaders Should Do Next: Leading Beyond the Dark Enlightenment The challenge for business is not to resist AI; it is to re-introduce friction intelligently. Efficiency has had its run. The next competitive advantage will come from leaders who understand that speed without reflection is drift, and automation without context is risk. Audit efficiency versus learning. Ask what your organization has understood, not just what it has automated. If AI made the process faster but did not make the people smarter, you have optimized the wrong variable. Reward dissent. Friction is not failure; it is evidence of trust. Teams that can disagree openly are still thinking. Consensus without argument is a symptom of automation culture. Protect provenance. In the age of synthetic everything, authenticity is infrastructure. Know where your data originates, who touched it, and how its meaning has evolved. Without memory, truth decays, and when truth decays, judgment follows. Build with humility. Treat models as maps, not worlds. Ask what they cannot see, the unquantifiable factors that drive loyalty, creativity, and moral judgment. Those blind spots are where your competitive advantage lives. AI will not end leadership; it will expose it. The Enlightenment gave us freedom through knowledge. The Dark Enlightenment offers knowledge without freedom. The leaders who thrive next will be the ones who can tell the difference and who choose to serve meaning over momentum. Toward a New Architecture of Doubt: Outgrowing the Dark Enlightenment The Dark Enlightenment is not a future to inherit; it is a symptom to outgrow. Progress without conscience becomes recursion. Governance without consent becomes optimization theater. Intelligence without humility becomes ideology. AI has exposed our era’s true faith: optimization as morality. We are watching institutions mistake automation for advancement, speed for insight, pattern for truth. The antidote is epistemic, not partisan: rebuilding doubt as a public good. Reality drift at scale means societies begin to model themselves through prediction rather than reflection. Policy, finance, and culture start reacting to what machines expect of us, not what is actually happening. The risk is not that AI lies; it is that it forgets there was ever a world outside its training data. If the Enlightenment made truth democratic, its dark twin is making it proprietary. Reality is no longer shared; it is licensed. The work ahead is to restore the architecture of meaning, to make truth not frictionless but resilient. Because what is at stake is not just the future of work or politics; it is whether reality itself remains a public space or becomes another private algorithm. The Dark Enlightenment tempts us with the illusion that optimization is progress and that intelligence can exist without wisdom. Resisting it does not mean rejecting technology; it means rebuilding the moral and creative friction that keeps intelligence human. Because in the end, the next Enlightenment will not be engineered, it will be chosen.
