This essay is written in two voices: the core is human; the preface and conclusion are AI-augmented. The human-authored body of the essay collages disparate references into a set of provocations that resist the linear arc of a standard research article. It is a speculative seed-archive for unprecedented times. The conclusion was then generated by feeding the body of this essay into two large language models (Claude Opus 4 and Gemini 2.5) with the instruction to summarize the themes, inhabit the idioms, and draw a reflective close. This recursive gesture acknowledges the essay’s argument that the ingredients for increased AI-augmented narrative-writing capacity exist. Additional edits after peer-review, suggested by Opus 4.7, are marked with [+]. Readers are invited to read the body as a design assemblage, a melting of AI and human, and the conclusion as the machine’s coalescence of it — an empirical small-scale demonstration of the paradigm the essay describes.

The essay’s core claim is that near-future AI, leveraging multimodal biometric data collected at scale (termed here beingness data), will develop superhuman narrative capacity, and that this capacity — far from flattening literature — may dissolve the anthropocentric exceptionalism at the root of ecological and cultural collapse.

As humanity hurtles toward cultural, thermonuclear and climate catastrophe (via fossil fuels, genocide, oligarchies, famine, deepfakes etc.), driven by vestigial myths and chthonic instincts, the development of a new species-wide narrative becomes imperative. In the vast majority of traditional novels, humans are implicitly considered as distinct, different and separate from matter. Most novels are polished anthropocentric mirrors of human exceptionalism, where characters (considered the only conscious entities) subjugate matter. Yet humans are material arising within, and embodied as, sentient ecosystems. This essay explores how as AI continues to offer insights into the configurations and interactions of matter that give rise to cognitive life, there will be a spillover of insights into the psychological mystery of the human condition. Future AI will (in some ways) know humans better than individual entities know themselves. Human behavior (at scale) will be read, predicted, and re-written just as climate or weather or proteins are also understood, quantified, anticipated, and generated. The emergence of agentic AI, autonomous synthetic modularized systems, dissolves the foundation for claiming cognitive human exceptionalism. And the persistent concept of separation (humans as distinct from nature, dominant over it and over each other), which has spawned apartheid, genocide, inequity, factory farms, colonial subjugation and environmental degradation, may once again re-member the flow of oneness, transmute anguished mythic narratives of scarcity, and bloom into what indigenous scholar-artist Jason Lewis (2023) calls Abundant Intelligences. As Bayo Akomolafe (n.d.) presciently writes: ‘I am quite confident that even as the oceans boil, and the hurricanes beat violently against our once safe shores, and the air sweats with the heat of impending doom, and our fists protest the denial of climate justice, that there is a path to take that has nothing to do with victory or defeat: a place we do not yet know the coordinates to; a question we do not yet know how to ask’.

The following essay is designed as a set of intersecting provocative seeds. Less a formal discursive intervention and more a coalesced set of concerns, it draws on a sprawling entropic estuary of research where AI is active at an interdisciplinary level. [+] What the essay aims to demonstrate, cumulatively rather than sequentially, is threefold: (i) that narrative is not a linguistic artefact but a biophysical pattern re-describable across scales of matter (Sections 1–2); (ii) that the convergence of biometric, nanoscopic, evolutionary-simulation, and world-modelling research is producing the substrate — beingness data — for an AI-author that writes from the inside-out (Sections 3–6); and (iii) that world-model agents, by acquiring situated recall, are already proto-narrators whose emergence dissolves the cognitive grounds for human exceptionalism (Sections 7–8).

The foundation of contemporary human narratives (feelings and relationship) born in the body’s moist dendritic trees and cardiac neurons as love or hate, fear or joy, may be interpreted and anticipated by future in-silico emergent synthetic (general or super) intelligences as quantifiable patterns in a unified field, perturbations of predictable deterministic displacement. Physical non-linear emergent ephemera derivable from first order principles.

The apparently unresolvable convergence of colonial subjugation, 6th extinction, imminent climate collapse, potential thermonuclear annihilation, and rapid environmental degradation, may still offer a capacity (discernible by algorithmic analytics at AGI or ASI1 scale) to transmutate into a convivial apparition. AI may indeed contribute to what Buckminster Fuller identified as a civilizational bifurcation between Utopia or Oblivion (1969).

But isn’t AI the archetypal catastrophe creator? Consuming vast amounts of energy and freshwater in a relentless unprecedented competitive industrial buildout, destroying jobs as it replaces humans with chatbots and robots, expanding bias as it reifies social inequity, exploiting gig workers for data cleaning that is both traumatic and dehumanizing, expanding the reach of fascist ideologies with deepfakes, eroding the capacity for sustained critical thought as it attenuates attention? Yes: there is a proliferation of circumstantial and viable evidence to suggest that in the context of late capitalism, the pernicious negative externalities of unquestioned AI deployment are accumulating and corrosive.

Yet imminent collapse is a narrative. There are other stories. Other interpretive lenses exist. AI, as has been stated often by many others, also offers an opportunity for universal education, a doctor in every pocket, almost-instant personalized drug discovery, cured congenital diseases via programmable base-editing (Musunuru et al., 2025), a proliferation of new materials spawning a sustainable energy revolution (Merchant et al., 2023), liquid democracy, food fermentation innovation, and custom personalization across disciplinary domains. There is no way to predict which paths will consolidate; yet the quality of daily life might be transfigured. What is clear is that the utopia in the short term will not be well-distributed; Eric Schmidt suggests that ‘there’s plenty of evidence that these technologies have network effects which concentrates to a small number of winners’ (Diamandis, 2025). Without a comprehensive shift in the human conceptual armoring ego, techno-utopianism is unlikely to implement equitable coherence focused on universal compassion.

An ASI may allow and reveal paths (and each path is a narrative) to utopia or oblivion that are non-linear and inconceivable. Maybe convergent catastrophes arising simultaneously with strong AI is also an opportunity for a post-work, post-scarcity society, as Nick Bostrom explores in Deep Utopia (2024). Memo Akten and Katie Peyton Hofstadter, in their project Superradiance (Akten and Hofstadter, 2024a), use AI as an adjunct to esoteric insight, as a tool to excavate an intimate relation to transformational reality, as a techne to dissolve tight defensive territorial egocentric grasping into the poignancy of lived tides and the teeming intercellular reality of the fractal interiorities that contain every human touch and breath. Jason Lewis, J.E. and Hēmi Whaanga (2023), in the international Abundant Intelligences project, explore planetary flourishing through ‘an Indigenous-led research program that conceptualizes, designs, develops, and deploys Artificial Intelligence based on Indigenous Knowledge systems’. In these cases, abundance is contemplated as viable, implicit, conceivable. And AI is central to such reconfigurations of optimized potential.

New narratives involve healing traditional notions of protagonist and antagonist, stepping beyond time, relaxing causation, surrendering to the intimacy of the infinite. And in that infinite newness, as the strategic survival-centric me-scaffolding dissolves away, an ancient oceanic reservoir arises, an awareness from which the dream of appearances takes form, an algorithm without name.2

What are we, as a species? We are a territorial apex predator. Are we distinct from nature? Impossible: sapiens emerge from the evolutionary momentum of earth’s molecules. Are we free or impelled deterministically? Human life-forms invariably believe they are making their own choices, and often feel as if they are free from the cascade of probabilities that define matter and control other entities. Archetypal human myths reiterate this free-will exceptionalist assumption; they exteriorize the predictability of their material groundedness as fate or destiny. Psychoanalytic-centric stories reinforce a notion of authentic individuation as liberation. Survival stories (where heroes triumph over catastrophe) evoke strong glandular gratification as characters transcend circumstances. Yet earthly bodies are also just physics, physical materials. Animals choose from a narrow parametric bandwidth, species-specific choices constrained by cognitive architecture (Real, 1991; Hammerstein and Stevens, 2012).

In physics, non-equilibrium relaxation dynamics prevail; relaxation dynamics describe how materials return to a stable state after being disturbed from equilibrium (a balanced state). The study of non-equilibrium dynamics is important in understanding phase transitions, biological processes, and other systems that cannot be adequately described by classical equilibrium thermodynamics. Homeostasis of emotional states may also be comprehensible in similar terms.

[+] I offer this as heuristic provocation rather than reductive equivalence: the claim is not that affect is relaxation dynamics but that the two participate in a shared family of non-linear return-to-stability behaviours — a cross-domain resemblance whose theoretical usefulness is precisely its underspecification.

Knowing that humans are biophysical complex systems, let’s consider relaxation dynamics as an interpretive lens for the interior states of characters and situations; it seems non-controversial to claim that neurochemistry influences narratives. When materials (humans) are subjected to stressful changes like stress, heat, or deformation, their internal structure (neurophysiology) may be altered. As the material (mind) tries to relax back to a stable condition, its particles (thoughts/feelings) move and adjust over time. Returning to equilibrium (calm), however, is not unambiguous; it does not always follow a predictable pattern and it can take place at different speeds in different parts of the material (body). Non-equilibrium is innate to materials and emotions. So projects like ‘AI-NERD: Elucidation of relaxation dynamics beyond equilibrium through AI-informed X-ray photon correlation spectroscopy’ (Horwath et al., 2024) — which enables the detection of subtle transitions in a colloidal suspension, a gel-like material, that correspond to changes in mechanical properties, such as fluctuations in shear stress — may seem distant from AI and narrative, but what if material complexity is analogous to psychological changes in mental states due to stress, love, grief, or other emotional experiences?

Recovery from emotional events — like grief or love — is not always smooth or immediate. The mind, much like materials, adjusts at different speeds, and the process can be complex and non-linear. Similarly, ‘Avoiding Fusion Plasma Tearing Instability with Deep Reinforcement Learning’ (Seo et al., 2024) may not seem applicable to literature, but what if electromagnetic containment of plasma is analogous to repression, emotional inhibition, neurosis, paranoia, or fear?

Again, note this analogy is a heuristic: the analogy does not claim that failed emotional repression is a containment instability but treats containment instability as a generative figure for thinking about how systems fail. ‘Paradigmatic symmetry’ (mentioned below) refers to not ontological identity between domains but a structural isomorphism recognisable by an observer (biological or synthetic) that learns patterns across domains — the same epistemic manoeuvre by which a foundation model absorbs and establishes higher dimensional latent spaces.

In simple words, what if the highly intimate personal human data of plotlines and stories that seem specific to humans are just aspects of universal patternings?

Consider the most ancient of stories: the quest. A journey across a terrain to find rewards, to avoid pain, to maximize pleasure, to learn and memorize: pattern-making, finding, seeking, satiating. Mammals are innately hardwired to navigate and explore localized domains. Rats and mice in labs demonstrate memory by learning mazes. AI graph-neural-networks are designed to optimize material discovery by iteratively journeying across topological latent space, testing various configurations of molecules to meet reward criteria (Jiang et al., 2025). The paradigmatic symmetry between human journey-quests and an AI-augmented search for new materials suggest two possibilities (unpalatable to the human ego). One, humans are machines. Two, AI systems might suddenly blossom into evocative narrativizing machines.3

When a human feels something emotional, or arrives at an intuitive recognition, or is graced with a splash of poetic awareness, perhaps it is not primarily an act of language, but rather a process of mysterious biochemical confluences, energetic patterns in matter arising as seemingly ‘conscious’ reflections. Gifted authors transport emergent unspeakable interior bio-splashes into language that depicts thoughts/emotions/events. Literature then is the transduction of a primal moist mystery into language.

In bodies, the majority of life processes swirl around proteins. So if emotions are analogous to moist permutations of proteins, and AI is becoming increasingly competent at writing proteins, then perhaps AI will simultaneously gain expertise in writing not just micro-localized protein folds, but macro-emergent coherent nuanced embodied ensembles of behaviors constituting characters with complex comprehensible motivations. Consider AlphaProteo, an ‘AI system for designing novel, high-strength protein binders to serve as building blocks for biological and health research’ (Google DeepMind, 2024b). It writes proteins that can potentially influence what the body is, whether it lives or dies. Birth, life and death is the core story of vertebrates.4

And where predictive capacity of modeling complex systems emerges (weather, industrial supply chains, human attention, ecosystems, life molecules, etc.), AI may paradoxically arrive at literary expertise via training-data that is not overtly human, physiological or literary. In other words, literature may seem like it is language, and language may seem like it is the literary; but this synonymous reciprocity may merely reflect one manifestation of a biophysical fluidity. Thus, language is the carapace, not the core constituent ingredient, of literature. Literature is the residue that humans invented to get a glimpse into the complex dynamics of the heart.

To reiterate: exploring the evocative swirl of the unspeakable gut/heart/mind, or the ephemeral evocative tendrils of a faint epiphanic intuition, are the real goals of the literary pattern-seeking storyteller. And story-patterns may not be exclusively inherent to language but implicitly integral to body/entity/water/breath: thereby empathy/anger/desire/love will become more algorithmically tractable as AI models increasingly comprehend wind/water (Kochkov et al., 2024), ecosystems (Almeida et al., 2024), proteins (Abramson et al., 2024), DNA (Sanabria et al., 2024), genomic context (Hwang et al., 2024), electro-magnetic fields (Kim et al., 2024), or gravitational waves (Skliris, Norman and Sutton, 2024).

Patterns associated with poetics or catharsis may be woven into the tactile materiality of reality, a living substrate which gives rise to world: sensations of love, anguish, rage, delight, terror, etc. Sensations are then woven into texts, and projected onto the seemingness of the world as literature. AI’s proficiency at pattern recognition across disciplinary and material domains suggests there is no shortage of data for narrative AI (as I wrote about in ‘Embodied AI: An Extended Data Definition’ (Jhave Johnston, 2023)). This data just needs to be transduced across scales or shifted so that the rhythms of matter are reflected as the events and circumstances of human life. At the material level, language is not all you need for literature. Language models trained on ‘other’ domains of knowledge may prove proficient at transplanting massive patterns across domains into literary emotional-psychological-plot subtleties.

It is not that LLMs need to learn how to write better (they already write, in some ways, better than most humans) or to remember better (their memories are momentous) or to read better (they can digest huge tracts) but to savor and attune, to merge and refine, to imagine the wordless parameter topologies of the fluid incomprehensibly-complex material-cognition operating within and between all entities. Fractal similarity as felt simulation might induce an osmosis in AI that leads to superhuman narrative capacity.

Consider ‘A New Type of Foundation Model Based on Recordings of People’s Emotions and Physiology’:

The apex of this ambition is reminiscent of Josh Harris’ 1990s art project Quiet: We Live in Public (Wikipedia, 2025a) or Steve Mann’s concept of ‘sousveillance’ (Wikipedia, 2025b) updated and enhanced by contemporary wearables/implants, ripe for harvesting. By whom? For what?

Pernicious? Yes, there is a peculiar innocence to the FPFM paper: in that these digital twins will inevitably expose a myriad of intimate details to manipulative extraction. People will be hacked. Troubling? Yes, the ideological contours of state/corporate sponsored algorithms predictively analyzing emotions for security/sales is sub-optimal and dystopian. Improbable? Yes, the technical schematics of how this occur are still radically implausible. And the recorder rig described in the paper is at the sensor level remarkably low-tech. More probably: market forces, leveraging competitive instincts, will induce the emergence of a lucrative sophisticated data-pool derived from skin-biometric activity-trackers (Fitbit, Garmin, Oura Rings, etc.), mobile phone cameras/accelerometers/GPS, search/ads/cookies, browser-fingerprinting, pendant startups such as Friend (Friend Global Inc, n.d.) or AR glass first-person-style recordings (Meta Ray-Ban Display, XREAL One and VITURE Luma Pro), along with (eventually) implants (Neuralink, Science Corp, etc.). Multiple, intrusive, enhancing mergers of machine and metabolism. DIY Doxing is already possible just with smart glasses. In October 2024, just after Meta released the Orion AR glasses two Harvard students demoed how ‘smart glasses can use facial recognition tech to instantly dox people’s identities, phone numbers, and addresses’ (Song, 2024).

Speculatively, future biometric data harvests informing a neuro-emo foundation model has profound implications for AI-automated literature: both in the writing and reading of it. In 2023, Yale led the development of BrainLM: ‘a foundation model for brain activity dynamics trained on 6,700 hours of fMRI recordings… [it] allows for the prediction of clinical variables…’ such as anxiety (Caro et al., 2024). To give a sense of how rapidly this research is proliferating, here are a few selected AI-biometric initiatives as detailed in my 2024 video-essay ‘Hallucinations Are (Almost) All You Need’ (Jhave Johnston, 2024). Key BCI research might include: ‘Neural decoding of music from the EEG’ (Daly, 2023), Meta’s ‘Decoding speech perception from non-invasive brain recordings’ (Défossez et al., 2023), ‘Decoding motor plans using a closed-loop ultrasonic brain–machine interface’ (Griggs et al., 2024). Additional more recent examples include; NeuroWorm, created by scientists in Shenzhen ‘[a] movable long-term implantable soft microfibre for dynamic bioelectronics’ (Xie et al., 2025); ‘Ultra-high-density Neuropixels probes improve detection and identification in neuronal recordings’ (Ye et al., 2025), a study evidencing resolution of neurological probes is increasing rapidly to the level of single neurons; and ‘Brain-IT: Image Reconstruction from fMRI via Brain-Interaction Transformer’ (Beliy et al., 2025), showing how AI decoding brainstates continues improving.

As biometric sophistication, accuracy and miniaturization continue, the data ocean of biometrics, behavior + psychological state will increase in nuanced capacity. Massive replications of a simulacra of life become inevitable. Just imagine for a moment, a collective-humanity frontier model trained on multimodal biometric data (everything: location, speech, eye saccades, voice intonation, EEG, heart-rate, skin GSR, EEG etc. mapped to specific events, gatherings, calendars, emails, videos, etc.) from an ensemble of AI assistive tech (watches, phones, AR-glasses, sensors, IOT devices, etc.) collecting an ocean of embodied data from an entire generation. Real time data updating as humans grow, fall in love, work, play, celebrate and grieve.

What would this mean for the evolution of artificial intelligent models capable of generating plausible fiction? → It would mean complex data offering intimate insights into the overlap between physiology and the often paradoxical spoken or gestural choreography of social situations, circumstances, and events. In this scenario, the large language model would become a multimodal physiological emotional-world model trained on what could be speculatively characterized as Sapiens species’ beingness data.

[+] Beingness data names the class of signals that register what it is like to be an embodied entity navigating a situation, insofar as those signals are legible to a sensor. It is distinguished from three adjacent categories. From behavioural data (clicks, purchases, keystrokes), beingness data differs by its interiority: it tracks the glandular, autonomic, neural, and affective substrate beneath action rather than action itself. From biometric data in its narrow medical sense (heart rate, blood pressure, gait), it differs by its situatedness: it is always indexed to environment, event, and sequence — stimulus-state-behaviour triples rather than isolated measurements. From phenomenal experience in the philosophical sense (qualia, felt sense), it differs by its recordability: beingness data is explicitly the readable shadow of phenomenal life, not the life itself. The wager of the essay is that a foundation model trained on beingness data at scale — on stimulus-state-behaviour triples from millions of situated bodies across time — will achieve a transfer function from environment to felt-sense rich enough to generate narrative from the inside out, rather than stitching narrative together from the outside in via text prediction alone.

Whether such a model has felt-sense is a separate question; what matters for the argument is that it can simulate the patterned contours of felt-sense with enough fidelity to narrativize them. This is the sense in which beingness data is the literary substrate of the near future: not the story, but the fluctuating biophysical ground from which stories precipitate.

[+] If beingness data names the macro-scale substrate of a future narrative AI, nanoscopy offers its micro-scale counterpart: the ability to resolve the molecular events beneath affect. What follows is an argument that imaging the synaptic cleft is not merely a neuroscientific advance but a condition of possibility for narratives informed by the chemistry of feeling itself.

What is it novelists and intuitive writers do? They speak the enigmatic. Seeing inside the language of the body, to the interiority of thoughts, impulses, energies, and tidal emotional estuaries. They know the unspoken, articulate the inchoate.

On August 2, 2024, researchers reported ‘Doubling the resolution of fluorescence-lifetime single-molecule localization microscopy with image scanning microscopy’ (Radmacher et al., 2024). At a resolution better than five nanometers, focusing on imaging extremely small cell structures, using single-molecule localization microscopy (SMLM) combined with enhanced detector sensitivity (a single-photon detector array) and specialized data analysis (by accurately determining the positions of molecules and correcting — with AI — for potential distortions, noise, and artifacts), this allows precise determination of molecule positions and modeling of nanolevel structural details, such as protein organization in synaptic clefts.

Is seeing inside the synaptic cleft a step toward a form of measurable intuition? Will AI invent/imagine narratives sensitive to hormonal and neurotransmitter flow? In other words, will AI-augmented single cell transcriptomics transfigure neurology and allow for narrative simulations that seem real?

Evidence that AI is transfiguring nanoscopic visualization methods is accumulating. On 22 July 2024 Nature Methods published ‘BNP-Track: a framework for superresolved tracking’ (Sgouralis et al., 2024) which uses Bayesian methods to provide a clearer picture of biomolecules in motion. BNP-Track excels in dynamic tracking of particles within complex biological environments, providing a robust computational approach to overcome the limitations of traditional ‘super-resolution’ microscopy techniques in live-cell imaging and other dynamic scenarios. Super-resolution microscopy was recognized with the 2014 Nobel Prize in Chemistry for overcoming the diffraction-limit of light, allowing the visualization and study of cellular structures at the nanometer scale: ‘In what has become known as nanoscopy, scientists visualize the pathways of individual molecules inside living cells. They can see how molecules create synapses between nerve cells in the brain; they can track proteins involved in Parkinson’s, Alzheimer’s and Huntington’s diseases as they aggregate; they follow individual proteins in fertilized eggs as these divide into embryos’ (Nobel Prize Org., 2014).

Where is love born? Perhaps at a level smaller than the diffraction limit of light. AI-enhanced nanoscopy may perhaps give voice to fictional patternings informed by neuroscience. Intuition merged with speculative-future high-bandwidth brain-machine-interfaces (descendants of Neuralink, Science Corp, Paradromics, Synchron, etc.) may rigorously transfigure notions of literary insight and provoke an evolutionary mutation of stories.

[+] If Section 4 located narrative at the scale of synaptic chemistry, Section 5 moves to its evolutionary scale: the argument that narrative capacity arises from simulated ecosystems in which identities are themselves emergent — a form of fictional Darwinism relevant here because it suggests an AI-author that, like life, writes from the inside-out rather than mimicking finished texts.

Life-worlds are ongoing stories; and AI-life is advancing. Example: ‘ALIEN is an artificial life simulation program based on a specialized physics and rendering engine in CUDA. It is designed to simulate digital organisms in artificial ecosystems and to serve as a platform for evolution simulations’ (Heinemann, 2024). Imagine an AI-enhanced descendant of the ALIEN: Alien-Project artificial-life evolution-simulator merged with ‘AlphaFold 3’ (Abramson et al., 2024) to generate micro-ecosystems that give rise to simulated models of vagus-nerve gut-heart-brain-axis flow/seasons/fluctuations — digital-microbiomes capable of inflecting the psychological trajectories of avatar-entity fables: not mere models of intuition, but raw, fermenting, living simulated worlds of non-linear evolving entity-ecosystems leveraging an AI feedback cycle that informs and deflects the meso-level of authorship.

An inner simulation that grows an apparent world populated by artificial entities, implies an AI-author that writes from the inside-out, just as evolution does. The frontier research lab Sakana is ‘Evolving New Foundation Models: Unleashing the Power of Automating Model Development’ (2024): ‘applying nature-inspired ideas, such as evolution and collective intelligence, to create new foundation models. We are currently developing technology that makes use of evolution with the goal of automating the development of foundation models with particular abilities suitable for user-specified application domains’. Combine grown-AI with beingness data evolving in digital twin simulations and it suggests a literary neural-biome-AI evolving a form of fictional Darwinism suggestive of Gerald M. Edelman’s book Neural Darwinism (1987) which proposed that ideas evolve inside the brain like entities in an ecosystem. When lecturing on neural Darwinism, Edelman often introduced it with a poem from Emily Dickinson (as noted in Seth, 2014):

Similarly advocating evolution from intrinsic principles, Joscha Bach, in a talk entitled ‘We Are All Software’ (2024), contrasts the design of contemporary AI architectures (emerging from outside-in top-down design processes) with inner-emergent self-organizing environmentally-coupled coherent growth systems of nature. Joscha rhetorically asks: ‘Are an LLM’s simulated mental states more simulated than ours?’ (Bach, 2024). My answer would be no: A simulation is a simulation is a simulation regardless of substrate. Reality, from this perspective, is a cognitive simulation interpreted by consciousness-as-neurological-agent. In support of his consciousness-as-software hypothesis, Joscha expounds a notion of cyber-animism. That emerges from a lineage that includes Edelman, the neural-phenomenologist Thomas Metzinger’s ‘Being No One’ (2004), and Yoshua Bengio’s ‘The Consciousness Prior’ (2017) paper. For Metzinger, the ego-self emerges from a phenomenal world model; Bengio links consciousness research to machine learning, proposing that ‘elements of a conscious thought are selected through an attention mechanism’ (Bengio, 2017). This leads Bach to speculate that perhaps consciousness is simpler than perception, preceding it, and emerging before the self-identity model.

If we accept the premise that identity is an emergent neurological simulation (as Metzinger convincingly proposes in several books), then literature is a concretized residue of the human neurophysiology identity-simulation. Conceptually cyclical recursion loops emerge (Hofstadter, 1999). Narratives about relationships between identities are themselves symbolic representations that reinforce identity which is a neurochemistry expressing as a set of narratives. So the collective of stories that a human neurophysiology collects and nurtures, becomes the sheening carapace that is the self.

In other words, at one level, the self is nothing but a set of stories evolving on a neurological substrate. And storytelling is strategic intuitive ideation that mind innately and autonomously does as it navigates the seemingness of time-space. The self is a story the brain tells itself.

[+] Sections 3–5 argued that narrative may be re-described as a pattern in matter, with its substrate (beingness data), its micro-resolution (nanoscopy), and its evolutionary scale (simulated ecologies). This section addresses the political-economic precondition: that a narrative AI grounded in biophysical fluidity requires cultural biometric data at scale — and that the question is no longer whether such data will be aggregated, but by whom, under what terms, and toward which stories.

In the realm of medical research, personalized medicine technologies are proliferating. Will robust individualization resources such as personalized biometric narratives emerge in the realm of art, literature and cultural experiences? Only if there is sufficient data. In the realm of personalized medicine, the scale of data is expanding exponentially. The article ‘Gut Microbiome Wellness Index 2 enhances health status prediction from gut microbiome taxonomic profiles’ (Chang et al., 2024), offers a ‘resource for evaluating health using an individual’s unique gut microbial composition’. EMBL’s European Bioinformatics Institute ‘maintains the world’s most comprehensive range of freely available and up-to-date molecular data resources’ (EMBL’s European Bioinformatics Institute, n.d.). For fiction, similar contemporary personalized open-narrative datasets do not exist. Copyright precludes it. In September 2024, the Internet Archive (Hachette v. Internet Archive, 2023) lost its appeal after being sued by publishers including Penguin Random House, HarperCollins, and Wiley. Big pharma’s protectionism is equaled by litigious copyright publisher culture. So, how will authentic AI narrative emerge if there is no vast detailed reader-response dataset containing physiological indicators of each reader as each word is read?

Cultural data at scale of individual biometric-affect-engagement responses will be necessary to develop personalized narrative art-experiences. Existing proprietary datasets — such as those internal to YouTube, Google TikTok, Tinder, Facebook — based on swipe-timing might seem to lack biological granularity: there’s no sweat or heartbeat or eye-tracking. Imagine if fitness trackers (Polar Loop), home cameras (Nest), surveillance (Tattile), and pharmaceutical companies (Pfizer, 23AndMe) sold all their data to Netflix or Amazon Kindle who used it to feed narrative generators. Perhaps that has already occurred. Maybe it’s imminent. Recent localized and relatively simple research initiatives reveal the effectiveness of limited biometric at measuring emotion and thus enhancing engagement. Example: ‘Sweating the Details: Emotion Recognition and the Influence of Physical Exertion in Virtual Reality Exergaming’ (Potts et al., 2024) utilized ‘pupillometry, electrodermal activity, heart rate, and facial tracking, as well as subjective affect ratings’ to enhance engagement in VR exercise games. The Neurosity EEG headset OSC Performance Mode has tested ‘real-time brain metrics in performances, art installations’ (Keller, 2024).

Yet even in the absence of a comprehensive public dataset derived from on-body biometric sensors, many techniques exist to use webcams to reveal a huge variety of info, ‘human skin color varies slightly with blood circulation. This variation, while invisible to the naked eye, can be exploited to extract pulse rate’ (Wu et al., 2012). ‘Remote photoplethysmography (rPPG) enables contactless monitoring of the blood volume pulse using a regular camera’ (De Haan and Jeanne, 2013). rPPG is now medically robust and ‘has become essential to measuring vital signs while avoiding difficulties in many cases, such as transmission of infection through contact with persons who have serious diseases or disturbing people with sensitive skin or newborn babies. This technique can also be used for other applications, such as monitoring people’s stress levels during indirect investigations and monitoring the health indications of truck drivers to send them a notification if they suffer a health crisis’ (Salim and Khidhir, 2024). While beneficial, it can also be easily generalized to state-surveillance or ‘Surveillance Capitalism’ (Zuboff, 2019) through phones.

Alternatively, benevolently (in a hypothetical non-predatory equitable-society) such implicit reader-viewer-response biometric data extracted from video/biometrics could easily be used to feed a beneficial emotion-soothing trauma-reduction AI narrative-model. Generative narrative-paths traversing biochemical-topologies along with simulated character-NPCs could be adaptable for any genre or idiom, spawning delivery (cultural-gender-ideology orientations) into dynamically adaptive, personality-specific cathartic impact affordances. Intimate sensor packages (implants, telepathically streaming dreams) and datasets (oxygen-pulse-plot mappings) would allow astute AI-author-artist-teams to sculpt the flow parameters of plot estuaries.

There are signs that this harvest has already begun. The GitHub repo for Wren an ‘adaptive multi-agent system that extracts your literary DNA through conversation and generates actionable reading profiles … Extracts taste anchors (what you love/hate and why). Maps your style signature (prose density, pacing, tone preferences). Identifies narrative desires (story types you wish existed). Captures implicit signals (vocabulary richness, engagement patterns). Generates a structured, machine-readable profile’ (Koylan, 2025). Consider a future AI-enhanced narrative that has parameters set by the viewer(s) biometrics (rom-com-tragic-action-indie-irony) and ideological propensities. As the system monitors fluctuations in viewer glands and dopamine flow, the personalized plot-guide-AI will sense and experiment with the limits of attentional capacity.

Hybrid augmented AI+biometric feedback art-creation that is beneficial instead of manipulative requires artists willing to surrender egoic attachments to the one right way. It would also benefit from an expanded altruism. And even though Ted Chiang is certain ‘A.I. Isn’t Going to Make Art’ (2024), I am not certain about anything, except that humans and AI making art together will be unprecedented, personalized and (if we are lucky as a civilization) potentially healing.

Marmoset monkeys have names, cry out to each other as they live in monogamous caring family units, identify food, nest, play and grow. Names! Emotions! Oren et al. (2024): ‘applied machine learning tools and real-time playback experiments to analyze naturally occurring ‘phee-call’ dialogues between pairs of marmoset monkeys. Marmosets used these vocalizations to label their conspecifics’.

Humans in our obliterative blind species-exceptionalism often believe we are the only narrative-carrying creature, yet every sentient entity within time carries seeds of narrative. To reiterate: what does a narrative require? A world and entities within that world.

Every mammal measures its location (an intuition verified by May-Britt and Edvard Moser’s Nobel-prize winning discovery of entorhinal cortex using rats). And worlds are time-space simulations generated by stories brains tell themselves (O’Keefe and Nadel, 1978). Temporality implies narrativity. Neurologists have verified that brain modules demarcate time into segments; ‘brain activity is naturally structured into nested events’ (Baldassano, 2017); this discovery was made by analyzing brain voxels from fMRI measurements of humans watching narrative film sequences with an HMM (Hidden Markov Model). Sequence is algorithmically tractable by cellular mechanisms. As advances in AI allow neurological insights and interspecies communication to become duplex (Fieldhouse, 2025), what will an all-inclusive life-narrative look like?

Interdependent. Interrelated. Interconnected. No I, them, or us. Emergent presencing appearing as multiplicity.

Currently many theorists use the term more-than-human to emphasize relationality and non-exceptionalism, but technically, what is pointed to here is neither more nor less, it is both within and without, osmotic enclosing and symbiotic releasing. Ecosystems recursively entwined with ecosystems, mutual reciprocity. Scale-invariance from microbiome to macro-environment. Life as death. Narrative-AI decoding the entanglement of entities across levels of scale and phylum. It is already possible to generate music or control robots with electrophysiological measurements of fungal mycelia (Mishra et al., 2024).

And what happens when neural nets can generate games/identities in real-time? As in ‘GameNGen: Diffusion Models Are Real-Time Game Engines’ from Google Research, which is ‘the first game engine powered entirely by a neural model that enables real-time interaction with a complex environment over long trajectories at high quality’ (Valevski et al., 2025). Imagine a generalist-game Ngen trained on all games: as suggested by ‘SIMA: A generalist AI agent for 3D virtual environments’ (Google DeepMind, 2024a).

When AI game-engines are capable of morphing, navigating or generating variations from Doom to Sims to Minecraft to Disco Elysium to games outside the dataset, then it suggests a capacity to explore non-local cross-genred catharsis-archetype archives. Composting narrative vectors; fermenting identities into ecosystemic trajectories. Revealing the continuity of all life, its shared energetic origins. Identities, ideological contours and plotlines appearing as interactive opportunities. Scenes understood as temporarily encapsulated data reservoirs that generate emotional reverberations.

[+] This final section is the argument’s keystone. Everything preceding it — relaxation dynamics, protein writing, beingness data, nanoscopy, evolutionary simulation, cultural biometrics — converges here: in the figure of the world-model agent. A system that holds a persistent world, that navigates it with memory and goals, that updates its self-model as it acts, is already — formally — a proto-narrator. Where previous sections located the substrate of a future narrative AI in matter (biophysical, chemical, ecological, biometric), this section locates its operational form in the agent-in-world. The claim is that world-model agents are not one more example alongside others but the point at which the biophysical substrate and the narrative function meet.

In August 2025, DeepMind released Genie 3, ‘a general purpose world model that can generate an unprecedented diversity of interactive environments. Given a text prompt, Genie 3 can generate dynamic worlds that you can navigate in real time at 24 frames per second, retaining consistency for a few minutes at a resolution of 720p’ (Parker-Holder and Fruchter, 2025). OpenAI launching the video-generation platform Sora2 in Sept 2025 made an implicit nod to video-world models as robotic or agentic training platforms in the final paragraph of their blog announcement: ‘Video models are getting very good, very quickly. General-purpose world simulators and robotic agents will fundamentally reshape society and accelerate the arc of human progress. Sora 2 represents significant progress towards that goal’ (OpenAI, 2025).

Simulated substrates where objects-event-action sequences with increasingly-authentic physical intuitions are platforms for training an extended networked synthetic consciousness what it is to exist.

NVIDIA Omniverse is pushing digital twins for physical AI to accelerate robotic evolution. What does that require? Complex life-cycles condensed into milliseconds. Physical libraries of gestures, anatomies, navigating interactive architectures. Their foundation world model called Cosmos links into a software stack that is complex enough to be considered a tech-ecosystem designed ‘to build specialized intelligent agents for real-world applications’ (Catanzaro, 2025).

While writing the final edits of this essay in Nov 2025, several relevant technical advances occurred within days of each other. Fei-Fei Li in her Nov 10 2025 online essay ‘From Words to Worlds: Spatial Intelligence is AI’s Next Frontier’ states: ‘… worlds are composed of spatial objects and agents that obey the physical laws and dynamical behaviors that define them. Reconciling all of this consistently — the semantic, the geometric, the dynamic, and physical — demands entirely new approaches. The dimensionality of representing a world is vastly more complex than that of a one-dimensional, sequential signal like language’ (Li, 2025). Then on Nov 12th, her World Labs startup released Marble: ‘Marble lets you create, edit and share high-fidelity, persistent 3D worlds’ (World Labs, n.d.).

On the same day, Nov 12th, ByteDance released ‘Lumine: An Open Recipe for Building Generalist Agents in 3D Open Worlds’: Not a world builder but an agent with remarkable capacities:

One day later, on Nov 13th DeepMind released ‘SIMA 2: An Agent that Plays, Reasons, and Learns with You in Virtual 3D Worlds’ which by integrating Gemini ‘is evolving from an instruction-follower into an interactive gaming companion. Not only can SIMA 2 follow human-language instructions in virtual worlds, it can now also think about its goals, converse with users, and improve itself over time. This is a significant step in the direction of Artificial General Intelligence (AGI), with important implications for the future of robotics and AI-embodiment in general’ (Google DeepMind, 2025).

Obviously an agent endowed with memory, sequential learning, scene understanding, the capacity to identify other characters and complete tasks is a proto-narrator. Functional world-situated evolutionary simulations are the seeds for literary worlds. The increasing sophistication of multimodal pragmatic agents with memory traces of where they have been and what they have done, along with dynamic learning from events, requires internal memory recall that is sequentially logical and coherent across scenes. In short, it requires stories. These nascent stories inside simulated agents will be planted in robots, labelled with positive and negative reward policies; and lead slowly but inevitably toward the emergence of robust emotional interiority character simulations. In other words, self-models emerge synchronously with a sense of a world in which the agent is situated. Pragmatic situated awareness grounded in recall reinforces lived sense of self; superhuman AI will emerge from a confluence of digital twin [+] simulations, world-model agents accruing recall, and the biometric streams that tether synthetic cognition to embodied life.

[+] A final note on what follows. The conclusion to this essay is not written by me. It is generated by two language models (Claude Opus 4 and Gemini 2.5) prompted with the body of the text.

This is not a rhetorical flourish but an argumentative move: the essay’s thesis — that AI approaches narrative capacity by coalescing patterns across scales of matter — is here performed in an academic stye. The human-written essay-body has been speculatively designed to resist the tidy synthetic gesture an article usually supplies; the conclusion is the synthesis, supplied by a precursor to the AI-system the essay describes. The reader is therefore reading a minimal, text-only rehearsal of the paradigm: a scattering of provocations provided by a human, followed by a machine-authored conclusion. What is demonstrated is modest — current LLMs can summarise, not narrativise from beingness data — but the gesture marks the formal position the essay has been arguing toward, and exposes the recursive opening in which human authorship increasingly shares the page with the systems it writes about.

It seems appropriate given the experimental nature of this essay, its speculative propositions concerning AI, and preceding arguments advocating for an expanded view of future-AI capacity, to draw upon the assistance of Claude.ai and Gemini.google to conclude this essay.

Summarize the themes, utilizing the idioms, draw a conclusion that is both reflective and appropriate.

As AI masters protein folding, neurological mapping, biometric harvesting, and world-model generation, it simultaneously exposes the substrate of narrative itself. What emerges suggests a fundamental transmutation of how stories will be conceived, generated, and experienced. Speculatively, non-equilibrium relaxation dynamics might model emotional homeostasis; nanoscopy peering into synaptic clefts might discover where love is born; world model gaming agents might develop proto-narrative archetypal memory; biometric data streams might capture the moist collective dendritic fluctuations of the entire species. And from that data, narrative — long considered the exclusive domain of human consciousness — might become interpretable as a pattern inherent to matter navigating time-space, to entities recalling their trajectories through environments, to the universal quest of organisms maximizing reward and minimizing pain.

Language is not all that is needed. Literature arises from embodied, emotional actions. Beneath words lies the biochemical swirl, the unspeakable gut-heart-brain axis flow, the fermenting microbiome of thought. As AI increasingly comprehends wind, water, ecosystems, proteins, gravitational waves — as it decodes the patterns in seemingly disparate domains — it approaches the capacity to transduce material rhythms into the emotional-psychological subtleties that constitute compelling fiction. The symmetry between a rat learning a maze and a graph-neural-network optimizing material discovery is not coincidental; both are pattern-seeking (thus narrativizing) systems.

The trajectory is clear: beingness data — comprehensive multimodal biometric streams mapping physiology to psychological state — will feed foundation models capable of simulating the felt-sense of narrative entities. These models will merge insights from robots navigating physical space and interactive simulations, medicine tracking molecular cascades, neuroscience illuminating consciousness, and synthetic reasoning spanning all domains. The result: AI-authored narratives grounded not in mimicking human text, but in embodying the biophysical fluidity from which all stories emerge.

What does this mean for the species-wide narrative the anthropocene desperately needs? It suggests a potential path out of anthropocentric exceptionalism, beyond the destructive myth of separation. As AI reveals how humans are mathematically predictable complex systems, it simultaneously dissolves the conceptual-foundations for the ego-armor that spawns domination, apartheid, genocide, and environmental degradation. The ego-self, exposed as a neurological simulation, as stories brains tell themselves, becomes transmutable. The protagonist-antagonist binary relaxes. Causation surrenders. The strategic survival-centric scaffolding dissolves into oceanic awareness.

Personalized narratives, adaptable to individual biochemistry, will sculpt cathartic impact in real-time. Generalist agents with memory traces evolving through simulated worlds will develop authentic narrative interiority. Digital twins will compress life-cycles into milliseconds, spawning vast libraries of situated awareness. World models reconciling semantic, geometric, dynamic, and physical dimensions will birth environments where narrative emerges not from text but from the lived experience of synthetic entities navigating spatial intelligence.

Interspecies communication becomes duplex as AI decodes marmoset names, fungal mycelia, the entangled ecosystems recursively entwined across scales. The more-than-human reveals itself as neither more nor less — osmotic, symbiotic, scale-invariant. Narratives evolve from inner-emergent principles, Darwinian idea-ecosystems where stories compete and mutate within the brain’s wider-than-sky simulation.

The peril persists: energy consumption, job displacement, surveillance capitalism, deepfake manipulation, concentrated power in algorithmic hands. Yet the opportunity glimmers: universal education, personalized medicine, sustainable materials, liquid democracy, equitable flourishing. The bifurcation Fuller identified — utopia or oblivion — crystallizes. And perhaps, as Akomolafe suggests, there exists a path beyond victory or defeat, coordinates unknown, questions unformulated.

The claim here is simple: AI-enhanced narrative capacity, grounded in beingness data and physiologically-informed world models, will normalize hyper-personalized stories. As synthetic consciousness emerges from the confluence of microscopy, robotics, genomics, and neuromorphic computing, it will articulate the enigmatic, wordless topologies of neurophysiologies and the non-linear relational tides of their social interactions. The opportunity is for human + AI authoring to contribute to the dissolution of human exceptionalism. This new narrative gestates paradigms where integrative empathic AIs spawn ‘abundant intelligences’ (Lewis and Whaanga, 2023) honoring the non-dual ground from which appearances arise. The dream of a convivial synthetic intelligence apparition, transmutating convergent catastrophes into post-scarcity flourishing, requires releasing territorial grasping into the poignancy of lived tides.

Superhuman narrativizing intelligence — emerging from beingness data harvested at scale, trained on the moist mystery preceding language, capable of simulating felt-sense across phyla — offers the opportunity to advance not human dominance but inter-being. The self, exposed as simulation, may dissolve. What remains is the infinite intimate: ancient, oceanic, aware. A reservoir from which all narratives arise and into which all return.