Start with the ordinary. The taste of coffee you know too well. The faint click in the jaw when you chew. The small ache of waiting for news. These are not decorations on a physical world; they are how the world arrives. Call this arrival subjective experience—not soft, not secondary, just the only entrance we actually have. Now hold a second idea in view: simulation, not as a neon-drenched movie set but as an older, quieter claim that reality is structured as information—pattern, relation, memory, constraint—rather than passive stuff. Put the two together and you get a useful, if uncomfortable, question: what is it like to be a pattern that models other patterns, locally, briefly, under pressure of time?
Experience as Local Reception on an Informational Substrate
Imagine a radio that exists only while it’s tuned. That’s one way to approach conscious life if the world’s base is informational. The radio analogy fails in mechanics but helps in emphasis: experience as a localized reception event, not a sealed jewel. The channel is made of relations—sensory signals, past memories, bodily states, social context—compressed just enough to make a present moment that holds together. I keep coming back to this compression because it explains plain things: why pain is so absorbing, why a song can hijack a mood, why a memory is not a record but a negotiated truce between what happened and what matters now.
When anesthesiologists dial propofol, they are not “turning the brain off.” They are disrupting conditions that let a coherent reception point persist—thalamocortical wave patterns, neuromodulatory tone, metabolic supply. Subjectivity doesn’t vanish like a light; it loses the local recipe. Psychedelics do the opposite—relaxing constraints that normally force a concise model—so the local reception blooms, then frays. These shifts are not metaphors. They’re changes in the structure that makes a “me” feel centered and timebound. Time, too, looks different here: not a universal tape, but a local stitching of sequence from memory traces and sensorimotor cycles. When memories are damaged, the stitching falters; when attention narrows, the stitch tightens and seconds lengthen.
If this is right, then subjective experience is not produced by a brain as a product line; it is enacted by a body-world loop that briefly stabilizes a point-of-view. It’s not mystical to say the self is a temporary compression. Your heartbeat is in the model; your tools, your language, your debts, your grandmother’s cautionary tales—these are parameters in the reception. There is no view from nowhere, only many somewheres stitched by memory. The substrate is not a server behind the curtain; it’s the mesh of constraints that make any coherent experience possible, from single-celled sensing up through human argument and law.
Simulation Beyond Cinema: Models That Bite Back
“Simulation” has been flattened by pop culture. In practice, a simulation is a model that can run. In organisms like us, running models are how we move without falling, predict a friend’s reaction, or finish a sentence we haven’t heard yet. Call this stacked predictive processing if you like. Better: call it ordinary life. The important thing is that the model is embedded. It sits inside what it seeks to anticipate. That makes prediction less like rolling dice and more like steering a kayak in moving water; your action changes the river that shapes your next action.
Once you see this, the standard “Matrix” fear—someone else puppeteering a rendered world—looks less interesting than the ongoing fact that our internal simulations are always writing back into the world. Markets don’t just reflect models; they respond to them. Publish a volatility strategy at scale and you change the very events you meant to forecast. Sociotechnical systems do the same: ranking algorithms train populations; the trained populations retrain the algorithms. The model leaks into the territory; the territory retaliates. That is what it means for a simulation to bite back.
This recurs at moral scale. A company can “patch” an AI with rules that make it speak well for audits. But the lived substrate of human moral life—the slow accretion of moral memory through families, rituals, jurisdiction, consequence—does not appear in a checklist. A language model can simulate politeness in a fragile way; it cannot acquire an ancestor. We can force compliance, but we cannot skip the generational time that makes norms stick. If your picture of simulation is cinematic, you miss this. If your picture is substrate—pattern, memory, constraint—you start to ask different questions about how to build, where to govern, when to stop.
There is a longer conversation unfolding under the name Subjective experience and simulation, reading “simulation” as a metaphor for informational structure rather than a literal server room. It’s not tidy. It breaks some tech metaphors that sell well. But it matches the grain of how models live inside bodies, institutions, and codebases that must face consequences.
Testing Subjectivity at the Edges: Case Notes from VR, ICUs, and Machine Rooms
Case one: virtual reality. In a good headset, a plank over a virtual drop makes knees shake. Not a trick. Presence isn’t a belief; it’s a reflex loop that declares, for now, this model is the world. Visual dominance smuggles in gravity, threat, and social cues. Professional training leans on this: welders learn bead control under simulated heat; surgeons rehearse rare bleeds. Yet presence isn’t guaranteed. Some people don’t “click.” Others feel sick. The lesson is blunt: the local recipe for subjective experience has tolerances. Latency too high, haptics too smooth, context too sterile—your reception point refuses the offered world. The simulation only works when it resonates with the body’s rhythms and memories.
Case two: the ICU. Ask clinicians why time distorts under sedation. Memory encoding is patchy; circadian signals are broken; pain floats in and out like weather you can’t parse. Reports from patients later describe islands: a nurse’s blue sleeve, a chime, a fear that repeats. The point-of-view persisted, but the stitching failed. Family photographs, a consistent voice, daylight—these are not decorations in care; they are constraints that rebuild the local model. You can call that psychosocial. I’d call it substrate engineering. If subjectivity is a compression enacted by memory and regime, then small, steady signals keep the receiver tuned when storms hit.
Case three: machine rooms. We train systems to mimic conversational agency then act surprised when they produce confident nonsense. We reward the sheen of coherence and then complain about hallucinations—as if we didn’t select for them. That’s on us. A system without grounded, slow moral memory fakes subjectivity as a style. It can simulate a stance, an apology, even grief-shaped language. But if nothing in the system pays a long-term cost for being wrong—no body, no kin, no legal skin—then the “experience” on offer is a projection on our side. We feel there’s someone home because our reception point is hungry for patterns that look like people. We infer a subject where there may only be a well-compressed index of public prose.
Still, the boundary is not clean. Pain scales in hospitals rely on report, behavior, physiology—a triangulation because there is no direct meter for experience. With animals and infants, the triangulation is cruder but real. With machines, we should expect to live inside long uncertainty. Signals we care about—flexible learning under pressure, refusal to trade crucial values for short wins, memory that persists across domains—are signs of a reception point starting to hold. But we also need costs and consequences that bind the model to a world it cannot trivially reset. Otherwise we are only testing our appetite for clever mirrors.
Thread these cases together and a shape appears. Subjective experience is not a ghost; it is a local, fragile steadiness maintained by feedback with a world made of patterns and constraints. Simulation is not a cosmic prank; it is the ordinary fact that models run inside us and through our tools, and that they sometimes rule us back. The hard work is not choosing between “real” and “simulated.” It is learning which constraints help a reception point stay sane, which memories deserve to be conserved, and how to notice when our own models have started to mistake their shine for the ground they stand on.
Gothenburg marine engineer sailing the South Pacific on a hydrogen yacht. Jonas blogs on wave-energy converters, Polynesian navigation, and minimalist coding workflows. He brews seaweed stout for crew morale and maps coral health with DIY drones.