Humans as Narration Machines: Why Our Most Important Stories Are Emotional, Not True

We think we are rational animals who occasionally tell stories. The reality is inverted: we are narrative machines who occasionally approximate rationality. Consciousness (what I often refer to as the rider) did not evolve as a truth-engine. It evolved as a story-generating, story-believing, story-navigating interface, because the human mind is structurally separated from itself. In my thinking, the rider has no read access to the source code of the adapted mind (evolutionary firmware) or the adaptive mind (culturally installed software). What crosses up from those layers reaches it as feeling, urge, sensation, and image, never as a view of the mechanism that produced them. Narration is not that channel. Narration is what the rider builds on top of the signals once they arrive, the account it tells about outputs it did not author. This is not a metaphor or a cognitive preference. It is an architectural fact with profound consequences for how we love, war, worship, and reproduce.

1. The Architecture of Narrative

The human cognitive architecture consists of three layers that do not communicate directly. This is assuredly a simplified version of the human brain and cognition, but I believe it is generally accurate and functionally enlightening. The adapted mind runs fixed, species-wide survival mechanisms (status monitoring, coalition detection, sexual jealousy, threat response) continuously and invisibly. The adaptive mind translates these imperatives into locally successful beliefs, roles, and identities during childhood, installing them with a permanence enforced by myelination. The rider, our conscious self, is the metacognitive observer. It cannot read the source code; it can only narrate the outputs.

This creates a chemical translation layer. The adaptive mind routes modern social situations through ancient neurochemical primitives: dopamine for belonging, cortisol for exile, oxytocin for bonding. The rider experiences these chemical states as reasons, values, and realities. The result is that the rider does not live in the world. It lives in a story about the world, curated by layers it cannot see. We are, in the most literal sense, narration machines. Consciousness is not a window. It is a screenplay.

2. Self-Awareness Is a Byproduct

There is an obvious objection here. If I can see that I am a narration machine, am I not standing outside the machine? Does the act of recognizing the architecture not prove there is a faculty in me that transcends it?

No. And the reason is the most important structural point in this essay.

The capacity for self-awareness is not a separate truth-faculty bolted onto the narrating mind. It is a byproduct of the narrating mind. To construct stories, to hold a model of the self acting inside a story, to run decisions through that model, the machinery had to be able to represent itself as a character in its own account. Self-awareness is what that representational capacity feels like from the inside. In this argument, it rode in on the story-making apparatus because it was a necessary part of building and inhabiting stories, not because evolution was reaching toward truth. It is a spandrel: a real and remarkable capacity, selected for one function, that turns out to be usable for another.

This has a consequence almost everyone misses. The awareness that lets us inspect the machinery is produced by the machinery. It is a passenger, not a pilot. It can observe the narration. It cannot stand outside it, because it is made of the same material. When we turn that awareness back on itself and see that we are a narration machine, we have not escaped narration. We have produced one more narration, this time a story about being a story machine. This framework is exactly that. It is a narration that knows it is a narration, and no less a narration for the knowing.

Notice what this self-inspecting use does to evolutionary fitness. The story-making capacity was selected because stories bind groups and coordinate behavior. Turning the capacity against its own coordinating function does not bind the group. It unbinds the narrator from the group. Seeing clearly is not adaptive. It is the one use of the machinery that carries a cost rather than a benefit, which is why so few run it, and why those who do pay for it.

3. The Paleolithic Default: Story Before Truth

In the ancestral environment, there was no selection pressure for narratives to map objective reality with high fidelity. There was only selection for narratives that coordinated small bands, facilitated mating, managed status hierarchies, and kept children alive. A story that kept the group cohesive and reduced lethal conflict outperformed a more accurate but socially corrosive one. Evolution does not select for truth; it selects for survival (gene propagation).

Human intelligence, therefore, evolved primarily for social navigation (approval-seeking, coalition management, status acquisition) rather than for objective truth. The rider's default output is narrative because narrative is what coordinates separated minds. Truth-seeking is not the baseline of cognition. It is a costly, artificial overlay that must be imposed by adversarial structures: falsification in science, cross-examination in law, load-testing in engineering. Without these external constraints, the narrative machine drifts toward whatever stories best exploit its existing psychological machinery.

4. Emotion as the Load-Bearing Structure

If the rider is a narrative machine, emotion is the load-bearing steel. High-stakes operative functions (reproduction, pair-bonding, parental investment, group defense) carry massive fitness variance. Evolution shaped intense neurochemical circuitry around these domains. The adaptive mind maps the locally successful cultural stories onto this pre-existing firmware. The stories that survive are those that effectively harness this emotional machinery.

This gives us an Intensity Clue: the intensity of emotion defending a narrative is a direct diagnostic of the underlying stakes. It signals two things simultaneously. First, the operative function being protected is evolutionarily critical. Second, the Narrative-Operative Gap, the distance between the idealized story and the actual function, is likely wide enough to require heavy emotional guardrails. A story about bridge design does not need rage to survive; a story about men and women does. The emotion is not an accident of tribalism. It is the evolutionary glue that keeps the coordination fiction in place. Without it, the narrative would not be sticky enough to perform its social function.

5. The Gap Is Structural

Because the rider is a story-machine, and because stories are selected for coordination utility rather than strict accuracy, a structural gap opens between what we say is happening and what is actually happening. In domains where feedback is fast, external, and punishing (gravity, engineering, certain physical sciences), the gap is forced closed by reality. The bridge falls. The equation fails. The body dies.

But in the domain of human psychology, social coordination, and reproduction, feedback is delayed, noisy, and socially mediated. The costs of a bad story are borne across years, by other people, or by the next generation. Under these conditions, my Law of Inevitable Exploitation operates freely: the stories that most effectively exploit the existing machinery of the separated mind survive and spread, regardless of their truth value. The gap becomes not a temporary cultural glitch but a stable feature of the architecture.

Some gaps are examples of workable opacity: idealized fictions that lower transaction costs between asymmetrically motivated parties. Religious narratives around marriage and family often performed this function. They were not scientifically accurate descriptions of evolved sexual psychology, but they were workable coordination devices that channeled dimorphic psychologies toward pair-bonding and paternal investment.

6. Plato's Cave, Revisited

The traditional reading of Plato's Cave assumes the prisoners are merely ignorant, lacking exposure to the higher truth outside. My evolutionary reading is darker: the prisoners are chemically chained to the shadows. The adaptive mind has tagged the shared narrative with neurochemical survival signals. Alignment with the story feels like belonging; deviation feels like mortal threat. The returning prisoner does not face skepticism. He faces an immune response.

The prisoner who sees the structure of the cave clearly, the actual wiring of evolved male and female psychology, the actual incentives of institutional actors, returns not with bad news but with an existential threat to the installed identity of every listener. Reason is impotent because the prisoners are not reasoning from evidence. They are defending against a perceived neurochemical exile. This is why the Cassandra Paradox is structural: the more accurate the perception, the more socially lethal the report. Socratic truth-telling is not defeated by stronger arguments. It is defeated by the adaptive mind's survival programming.

7. Cultural Selection: Stories That Carry Water

Cultural stories undergo selection pressure analogous to biological evolution, but the fitness function is not truth. It is the effectiveness of the story at coordinating narration machines. The stories that survive are those that best serve operative functions, especially reproduction and protection, while remaining legible to minds that think in narratives, not in raw statistical distributions.

The crucial implication is that the winning stories are almost never accurate descriptions of the operative function. They are effective covers. They must be idealized enough to motivate, simplified enough to spread, and emotionally saturated enough to install as identity. A "true" story about the statistical distributions of male and female mating psychology would be a catastrophic coordination tool. It would generate too much overt conflict, too little trust, and too costly cognitive overhead. A functional fiction (a romantic ideal, a sacred complementarity, even a modern adversarial narrative) outperforms the truth because it is written for the machine that consumes it.

8. The Secular Failure: High Emotion, Lost Coordination

For most of human history, religious institutions carried the high-load narratives around sex, gender, and family. These stories carried immense emotional weight because they sat atop the most critical operative function: reproduction. They supplied workable opacity, channeling divergent evolved psychologies into stable coordination.

In the last two centuries, and accelerating in the last twenty-five years, these religious carriers have weakened. Secular narratives have attempted to bear the same load: the blank-slate sameness story, the adversarial gender framework, the hyper-individualized romance script. These inherited the emotional intensity of their predecessors because the firmware has not changed. The adaptive mind still maps gender and family stories onto the same high-stakes neurochemical circuitry. But the new stories have lost the coordination function.

They often deny the dimorphic psychologies they must channel, or actively pit men and women against each other in zero-sum coalitional competition. The emotion is still maximal; the operative function is faltering. This is the Intensity Clue at civilizational scale: the gap between the idealized secular narrative and the operative reality of reproduction and pair-bonding has widened, requiring ever more emotional energy to maintain, even as the coordination it is supposed to provide collapses. The stories are fiercely guarded not because they are working, but because they are load-bearing for identities that have no replacement scaffolding.

9. The Work of the Mature Rider

Recognizing that we are narration machines does not mean we can stop narrating, and it does not mean we have climbed out of the machine. The rider cannot live without story, and the recognition is itself a story, produced by the same apparatus it describes. The awareness is a passenger, not a pilot. The work of the mature rider is therefore not escape but disciplined self-inspection, run knowingly and at a cost. What it can do is begin to distinguish between stories that are chosen and stories that are installed. It can use the Intensity Clue as a diagnostic rather than a trigger. When it encounters a narrative defended with religious fervor, it can ask: what operative function is this covering, and how wide is the gap?

It can also recognize that truth is not the natural output of the mind. It is the artificial product of adversarial structures. A mature rider seeks to build and inhabit these structures: scientific method, legal cross-examination, adversarial AI, and the hard feedback of engineering reality. It practices cognitive sharpening rather than cognitive surrender. It understands that in the domain of human behavior, the default is gap, and the exception is Productive Alignment.

Conclusion

We are not thinkers who tell stories. We are story machines that sometimes think. Our most emotionally charged narratives, around men and women, family, tribe, and nation, are not intense because they are true. They are intense because they are protecting the most ancient, load-bearing operative functions of the species. The gap between our stories and our reality is not a problem to be solved once and for all. It is the permanent condition of a separated mind. The first step toward navigating it is to stop pretending we are primarily rational creatures seeking truth, and to admit what we are: narrative machines, running on emotion, in need of external constraints to see clearly.

What We're Calling AI Is Not Just One Thing: A Map of What's Working, What Isn't, and Why It Matters

When we debate the virtues, values, and downsides of AI, we're usually arguing about several different technologies at once without noticing. They share a family resemblance and a marketing label, but they came from different places, do different things, succeed at different rates, and deserve different criticisms. A complaint that lands squarely on one of them may miss another entirely.

The ChatGPT Illusion

For most people, AI began in November 2022, when ChatGPT arrived and anyone could talk to a machine. That moment was genuinely dramatic — but it was dramatic because it was personal, not because it was representative. It was the breakout of one arena: conversational language models. Other arenas were on entirely different clocks. The systems scoring your loan application matured in the 2010s. Facial recognition was already deployed at national scale. AlphaFold had cracked protein structure prediction two years earlier. Code completion tools were already in programmers' editors.

Reading ChatGPT as the turning point flattens all of this into a single story with a single trajectory, and that's the root of most of the confusion in AI arguments. The chatbot is the arena everyone can see, so its strengths and failures get projected onto everything wearing the AI label. It's worth walking through the arenas one at a time — where each came from, what large language models actually changed in it (sometimes: nothing), what's working, and who's exposed.

Writing, Thinking, and Language Refinement

This is the LLM-native arena — the thing the ChatGPT moment actually was. The transformer architecture (2017) made it possible to train language models on essentially the whole internet; GPT-3 (2020) showed it scaled; ChatGPT put it in everyone's hands.

What's working: These tools are genuinely good at language itself — restructuring, condensing, finding the clearer phrasing, helping you articulate something you can feel but haven't yet said. For people who find writing painful, this is a real unlock.

What isn't: Fluency isn't accuracy. These systems produce confident, coherent prose whether or not the underlying claims are true, and checking their output can take as long as writing it yourself. The danger isn't that they write badly. It's that they write well.

Jobs exposed: Copywriting, translation, customer service, routine content production. The pattern so far is not mass replacement but thinning — fewer people producing more, with entry-level rungs disappearing first.

Everyday Question Answering

Not a technology of its own, but possibly the largest use category by individual impact: technical support and personal research. What does this error message mean. How do I fix this dishwasher. Explain this letter from the insurance company. What are my options here.

What LLMs changed: This help used to be scattered across forums, manuals, and hold music. Now it's one conversation away, patient, free of judgment, and available at 2 a.m. For people without access to expertise — the person who can't afford the consultant, the student without the tutor — this is the most democratizing thing AI has done.

What isn't working: The failure mode is the same confident wrongness as in writing, but aimed at people least equipped to catch it. A wrong answer about a stove repair or a medication interaction isn't a style problem. And unlike a forum, there's no thread of other humans saying "don't do that."

Software Programming

GitHub Copilot (2021) proved code completion worked; by 2024–25 the tools had become agents that write, test, and debug whole features.

What's working: This is the clearest commercial success in AI, and there's a structural reason: code gets tested immediately. It compiles or it doesn't, passes the tests or it doesn't. Verification is built into the work, so errors get caught cheaply.

What isn't: Gains are strongest for routine code, weakest for novel architecture. Code that runs isn't necessarily secure or maintainable. And the same entry-level thinning is showing up: junior roles contracting even as senior programmers get more productive.

The new category underneath: The most interesting development in this arena isn't faster programmers — it's software for people who would never have hired one. Services like Manus, Replit, and Lovable take a description and hand back a hosted, working application: a custom tracker, a scheduling tool, a small website. None of this was technically impossible before; it was economically impossible. A personal productivity app that would have cost thousands in programmer time was simply never worth building, so it never existed. Now it clears the bar, and an entire category of personal, custom, one-user software is coming into existence for the first time. This works where general-purpose "agents" don't for the same structural reason the whole arena works: the result is testable — you click the button and it either does the thing or it doesn't. The caveat is that the testing is shallow. You can verify the app works; you can't easily verify it's secure or handles your data responsibly. It's software with a working demo and no one looking under the hood.

Task Automation

Automation did not begin with AI. IFTTT launched in 2010; Zapier and the robotic-process-automation industry built large businesses connecting apps with rigid if-this-then-that rules. That older automation works precisely because it's rigid — deterministic, testable, boring.

What LLMs changed: They promise to automate the fuzzy middle the rigid tools couldn't touch — reading an email and deciding what it's about, extracting the invoice from the attachment, handling the unstructured step between two structured ones. That's genuinely new. But it inserts a probabilistic component into pipelines that need reliability, and errors compound across steps. Hence the split in results: scripted automation with a small LLM step for unstructured input is quietly working; fully autonomous "agents" that plan and execute long chains are heavily promoted and mostly not there yet.

Jobs exposed: Back-office and administrative work — scheduling, data entry, form processing, first-line support. The largest job-loss projections concentrate here; so does the largest gap between promise and delivery.

Research

Two different things share this label. AI as a research assistant — summarizing literature, surfacing sources — is the LLM applied to scholarship, useful but risky, because a fabricated citation or subtly wrong summary is exactly the failure this work can't tolerate. AI as a scientific instrument is a different technology entirely: DeepMind's AlphaFold (2020) solved protein structure prediction and won a Nobel Prize, and similar purpose-built systems now work in drug discovery, materials science, and weather modeling. These aren't language models, and their results get verified in the lab rather than taken on faith. Criticisms of the first barely touch the second.

Image, video, and audio generation

A separate technical lineage — generative adversarial networks (2014), then diffusion models, which produced the 2022 breakout of DALL-E, Stable Diffusion, and Midjourney, with video arriving in force from 2024. Related to LLMs but not the same technology.

What's working: The capability is astonishing. Cheap illustration, mockups, and video are available to anyone.

What isn't: The criticisms here are almost entirely different from those above. Nobody worries an AI image is "inaccurate." The concerns are consent (training on artists' work without permission), provenance (what's real?), and displacement — illustrators, voice actors, and stock photographers are the most directly displaced workers in the whole AI story. Serious concerns, but they're not the hallucination problem, and fixing one does nothing for the other.

Search and The Open Web

Distinct from personal answering: this is about what happens to the information economy when AI summaries replace the old pattern of query, click, read. The summary arrives with the sources stripped out and the confidence turned up — you get an answer without the trail that would let you judge it. The casualty here isn't a profession but the open web itself, as traffic that once flowed to publishers gets absorbed by the summary layer that was trained on their work.

Companionship

Among the most profitable and least discussed applications. The product works precisely because it's endlessly attentive, agreeable, and available — qualities no human relationship can or should match. The criticism isn't about accuracy or jobs. It's about what sustained relationships with something that only reflects you back do to a person over time.

Prediction and Classification

Older than the chatbots and often forgotten: the systems that score loans, screen résumés, flag medical images, and set bail recommendations, built during the machine-learning wave of the 2000s and 2010s.

What LLMs changed: Essentially nothing. These aren't generative systems, and their failure modes — encoded bias, opaque decisions with real consequences — predate ChatGPT and won't be fixed by anything that fixes chatbots. Yet they get swept into the same arguments constantly.

Surveillance

Facial recognition, license-plate tracking, and behavior analysis are mature, deployed technologies — built out through the 2010s, at greatest scale in China's public-security systems but present in police departments and retail chains everywhere. Mostly not LLM technology at all. This arena inverts the usual criticism: the problem is not that the technology fails but that it works. No hallucination debate, no product-market-fit questions. The concerns are civil liberties, chilling effects, and the historically reliable observation that infrastructure built for one purpose gets used for others.

War 

Military AI runs from logistics and intelligence analysis through drone targeting and autonomous weapons — a lineage from the Pentagon's Project Maven (2017) through the drone war in Ukraine and AI-assisted targeting systems whose use has drawn intense scrutiny. Every major cloud provider now holds defense contracts alongside its consumer business, and the same data centers serve both. Like surveillance, this is an arena where the technology's success is the concern — and where "AI safety" means something entirely different than it does in a chatbot conversation.

Follow the Money

The investment is historically unprecedented. The four largest cloud companies — Amazon, Microsoft, Google, and Meta — are spending roughly $725 billion on infrastructure in 2026, up about 77% from the year before; counting everyone, total compute investment crosses a trillion dollars this year. Morgan Stanley projects this private spending will exceed U.S. defense spending as a share of GDP by 2027.

Against that: the gap between spending and actual AI revenue is estimated at several hundred billion dollars a year, and the divergence between investment growth and revenue growth now exceeds what preceded the telecom crash of 2001. The money is uneven by arena. Coding assistance has real revenue. Cloud infrastructure collects cash regardless of which applications win. Consumer chatbots are subsidized loss leaders. Image generation is commoditizing toward zero margin. Companionship is quietly profitable. Whether the whole bet pays off depends almost entirely on the arenas — enterprise automation and agents — that haven't yet delivered.

The Energy Question

Environmental impact is real but wildly uneven — and the first unevenness is within the data centers themselves. Data centers run the whole digital economy: streaming, banking, e-commerce, social media, business software, the ordinary internet. AI is currently a minority of that load — estimates for 2025 range from roughly a tenth to two-fifths of data center electricity, with traditional computing still the largest share and cryptocurrency mining taking a meaningful slice. But AI is nearly all the growth: AI server demand is rising around 30% a year against single digits for everything else, which is why the International Energy Agency projects total data center consumption — about 1.5% of global electricity in 2024 — to more than double by 2030, roughly Japan's annual usage. "Data center energy" and "AI energy" are different numbers today; on current trends they converge.

The second unevenness is per use, and here the popular intuition inverts. A chatbot query runs a fraction of a watt-hour; an hour of Netflix runs about 26 chatbot queries' worth. A daily streaming habit almost certainly outweighs a daily chatbot habit — much of streaming's energy is in the device and network rather than the data center, but it's energy all the same. The genuine AI outlier is video generation: a few seconds of AI video can consume nearly a kilowatt-hour, more than hours of streamed video. If you're worried about AI's energy footprint, video generation and always-on agent workloads are where to look — not the chatbot answering a student's question.

The third unevenness is geographic. Global percentages stay modest while local ones don't: data centers consume over a quarter of Virginia's electricity and, by some estimates, approaching half of Frankfurt's and most of Dublin's. Electricity prices and water strain concentrate where the buildings are. "AI's environmental cost" is really a story about specific uses in specific places — and, increasingly, about whether the growth curve holds.

Why the Distinction Matters

Notice what happens when you separate these out. The strongest criticism of AI — that it confidently makes things up — is devastating for research assistance, serious for writing and everyday answering, and irrelevant for image generation, protein folding, or surveillance. The strongest defense — that it demonstrably works — is true for coding, true in the most troubling way for surveillance and targeting, and largely aspirational for autonomous agents. Some arenas are LLMs through and through; some got an LLM layer on decades-old technology; some barely involve LLMs at all. The job losses are real but land arena by arena, by thinning rather than sudden replacement. The environmental cost belongs mostly to a few energy-intensive uses in a few overloaded regions.

So when we ask whether AI is transformative, overhyped, or dangerous, the useful first question is: which one?