⊛

Holos: A Framework for Understanding Reality

Predictions16

In the version this site defends, Holos adds no new dynamical laws (the rules that say how things change over time) and changes none of the equations of physics. It is built on two ideas. First, a threshold, Φc\Phi_c. A system is an observer, with experience of its own, only once its parts are integrated enough to act as one whole. Second, Omega, the whole, with one experiencer living through every self. The commitments and tests below follow from established physics, from those two additions, and from the positions Holos takes on structure and mind (Axioms 1 and 4). Where a claim depends on one particular reading of quantum physics, the one without collapse, it is marked (Commitment 3 and Check C). The speculation at the end does not follow from any of this, and it is labeled as such.

The sections below separate three kinds of claims. The last section includes speculative extensions that aim to produce observable signatures, not just philosophy.

  • Commitments. What must be true if Holos is correct, independent of any future experiments.
  • Testability and its limits. What can never be tested (presence, the fact that experience is lived), two tests Holos could fail, one consistency check, and one standing bet about physics, with the fallback version of Holos stated in advance in case the bet is lost.
  • Speculation. Extensions that could follow under Holos on long timescales, stated with explicit alternatives rather than predictions.

1/3Commitments

What must be true if Holos is correct, independent of any future experiments.

For the definitions and the observer requirements, see Logic.

Commitments17

Commitments 1 and 2 are fundamental to Holos. A serious rival denies each, and if either is false, the framework fails. Commitment 3 belongs to Holos without collapse, the version this site defends. The version with collapse is declared in advance under Two versions.

1. Experience is local, and physics describes it only from outside

From Axioms 1, 3, 4, and 5, with the definitions in D7. See Axioms.

Two claims, and a serious rival denies each one.

Local. Experience occurs in no place except inside observers, systems past the threshold. It is not spread through matter or ambient in space or fields. Omega is the one experiencer, and it lives through observers. Whether the whole as such has any view of itself is not a question about a place, and Holos leaves it open. "Local" says where experience happens, not who has it. Structure without observers still exists, as pattern. Unobserved histories remain part of CC, the branches physics produces, and are simply never lived. (In the reading of quantum physics adopted here, every outcome a measurement could have occurs, each in its own branch, a complete, separate version of how the world goes.)

Only from outside. A physical description can be complete and still say nothing of whether anything is lived. The gap is not a missing fact. A perfect physical copy has the same inside. It is that the description is written from outside, like a floor plan that records every wall and still cannot say what living in the house is like.

Experience happens nowhere but in observers. Physics fixes all of it, but states it only from outside.

What this rules out:

  • Panpsychism: a flicker of experience in every particle, rock, or thermostat.
  • Illusionism: the view that experience does not exist, only the belief in it.
  • Strict physicalism of one kind: the view that physical language could, in principle, state that something is lived.
  • Dualism: experience as an extra ingredient beyond physics. The gap is in the description, not in the world.

1/5Local

Experience occurs in no place except inside observers, systems past the threshold. It is not spread through matter.

Anthropic principles (we should not be surprised to find ourselves in a universe that allows observers, since we could not find ourselves anywhere else) explain why observers find themselves in universes that allow them. They do not say where experience occurs, or why physics states it only from outside. Holos answers the first and names the second. Why there is experience at all, it leaves open.

Holos uses three words, each for one thing. Lived is where experience occurs. Lit is the causal past it draws on (everything close enough in space and time for light or signals to have reached an observer). And unlit is structure outside every observer's causal past. They are defined in Logic, D7. That "lived" means experienced by an observer is a definition, not a commitment.

2. Observerhood is thresholded

From Axiom 3. See Axioms.

1/5More is not one

Experience does not increase smoothly with more computation. Distributed processing can scale indefinitely without producing a single point of view.

Holos rejects the idea that experience increases smoothly with greater amounts of computation. Distributed processing can scale indefinitely without producing a single point of view.

What matters is integration. Well below the threshold, there is no unified internal state that could count as “what is happening for the system.” Well above it, in a system that meets the other observer requirements, experience is unavoidable.

Observerhood requires Φ≥Φc\text{Observerhood requires } \Phi \ge \Phi_c

In words, a system is an observer only if its integration, written Φ\Phi (phi), reaches the threshold, Φc\Phi_c. Φ\Phi measures how far a system's parts constrain one another as one whole, and the inequality is shorthand for being past the twilight described below. No one can yet compute Φ\Phi for a real brain, so the tests below use measurable stand-ins such as PCI.

Observerhood is neither ubiquitous nor optional. It appears when structural conditions for integration are met.

Whether a system crosses this threshold is a fact about how its parts are wired together, and integration alone is not enough. The integrated state must be about a world beyond the system (see the observer requirements in Logic). It belongs to the structural layer, the facts that hold the same for everyone, alongside the laws of physics, and it does not depend on who is looking. Observerhood is what qualifies a system to have a perspective at all. It is not relative to one. In the monist reading (the reading that the whole is the one experiencer), crossing the threshold is where a new self begins. No new experiencer does.

Two consequences follow. A physically identical copy of an observer cannot lack experience (Axiom 4). And the threshold is steep but not a mathematical line. There are clear cases on both sides, a narrow twilight between, fixed by structure, so a copy of a borderline system is borderline too (see The threshold in three claims).

3. Facts are relational but consistent

From Axioms 1 and 2. See Axioms.

Holos distinguishes two kinds of facts. Structural facts describe what is consistent, such as the laws of physics, the branches physics produces with their quantum weights, and whether a system meets the integration threshold. These are absolute and observer-independent. Registered facts describe what is lived, which outcome a system registers from its own perspective. These are always indexed to observing systems.

The relational commitment applies to registered facts. There is no absolute, observer-independent fact about which outcome is experienced. The structural layer, by contrast, is not relative. Without it, registration would have nothing stable to register.

This does not imply contradiction, and Holos is specific about why. No possibility is erased (Axiom 2), so observers never collide over a single shared outcome. Where registrations would be incompatible, they belong to different branches of the one quantum state, each internally consistent. There is no rule that the first observer fixes the truth for everyone. For far-apart events, relativity says observers moving differently can disagree about which came first, so there is no single “first” to appoint, and none is needed.

Within a branch, consistency is operational rather than abstract. Whenever two observers actually compare records, their records agree. Perspectives may differ while separated. Communication forces agreement. This is what "consistent" means in practice. Physics secures it. Observers who communicate are reached by the same signals.

Branches are not all equal. Each carries a weight, and those weights reproduce the probabilities quantum physics predicts (the Born rule). That is why, from inside, a 50/50 quantum experiment looks like a fair coin. Almost all of the weight lies with observers whose records follow the Born rule. The weights are structural facts, not amounts of experience, and for an observer they are odds of being in one branch rather than another. See Logic for the full account.

Registered facts are relative to observers. Structural facts are absolute. Observers who compare records agree.

1/3Structural facts

The laws of physics, the branches physics produces with their quantum weights, and whether a system meets the threshold. These are absolute and observer-independent.

Apparent collapse (the way a measurement seems to pick out one outcome) is therefore not a new physical process. Within each branch, records are already definite, a detector's click included. What an observer adds is not the definiteness but its being lived.

Everything that follows assumes these commitments. What comes next is what experiments can and cannot show about them.

Testability and Its Limits18

The central claim of Holos is that observation is where structure is lived, not a force. In the version this site defends, it changes no equation and moves nothing. But every experiment is a physical measurement, and an instrument only ever records physical change. So presence itself cannot be detected directly. An instrument that finds nothing extra is exactly what Holos predicts, because there is nothing extra to find. Presence is what the physics is like from the inside, not an additional signal beside it.

This is not a gap Holos has failed to close. It follows from the framework's own commitment, in the version defended here, that observation is dynamically inert. It has no physical effect, pushes nothing, and changes no outcome. The sharpest form of the objection is the unfolding argument (Doerig et al. 2019). Any system with feedback loops can in principle be copied by a loop-free one that behaves identically, so no behavioral test can tell which is conscious. It is aimed at integration-based theories like Holos. Holos accepts it for presence itself. The metaphysical claims (presence, and the whole it belongs to) cannot be settled by any experiment. Test A does not compare such twins. It asks, within real human brains, whether reports follow integration or behavior when the two come apart. The same limit is why artificial systems offer no test of their own. What a language model says about its experience is learned from human writing, so it is not evidence either way. Whether a machine crosses is judged by its structure (see Artificial Systems).

What remains testable is not presence but its structural preconditions: claims about what observation requires, and how registered facts behave. These live in the physical world and can genuinely fail. Two are stated below as tests, each with an explicit way to lose: Test A for the core, and Test B for the transition hypothesis. A third is stated as a consistency check. Its expected outcome is the one standard quantum mechanics already predicts, so it guards against contradiction rather than singling Holos out. A prediction Holos shares with rival theories cannot single it out, but a shared prediction it could fail is still worth more than one it cannot. Beneath them all sits a standing bet, stated after them, on the physics Holos adopts, with the version of Holos a loss would leave declared in advance.

1/3Nothing extra

An instrument only ever records physical change, so presence itself cannot be detected directly. An instrument that finds nothing extra is exactly what Holos predicts.

Test A: Consciousness tracks integration, not behavior18

Bedside assessment treats responsiveness as the sign of consciousness. A patient who follows commands is conscious, and one who does not is presumed not to be. No serious theory equates the two, but the proxy runs deep in practice. In Holos, what matters is integration, and integration can come apart from outward behavior. When the two diverge, Holos bets that experience follows integration.

1/5The bedside proxy

Bedside assessment treats responsiveness as the sign of consciousness. One who does not respond is presumed not conscious.

The bet is losable because cases where the two come apart already exist. People under anesthetic doses of ketamine, in REM sleep (the stage of most vivid dreaming), or in some seizures do not respond, yet later report vivid experience. Sleepwalkers, and people performing automatic actions during certain seizures (automatisms), move and respond, yet often report little. These are natural experiments that pull integration apart from responsiveness.

Objective

Across states where responsiveness and integration diverge, determine whether later reported experience tracks a measure of integration or tracks behavioral responsiveness and arousal (how awake the brain is, as distinct from what it experiences).

The protocol, fixed in advance

A test needs numbers, not intentions. Test A is frozen here, before any new data:

  1. Primary gauge: the Perturbational Complexity Index (PCI). A magnetic pulse is sent into the brain through the scalp (transcranial magnetic stimulation, TMS), and EEG records how the cortex echoes. A rich, widespread, varied echo scores high. A local or uniform one scores low. PCI captures integration together with differentiation, so it is a stand-in for integration, not a direct measure. The original PCI's published cutoff is 0.31 (Casarotto et al. 2016). A cortex-wide gauge.
  2. Second gauge: the same kind of measurement confined to the back of the brain's outer layer (posterior cortex), where dream reports have been found to track local activity. Its cutoff is set on calibration states (cases used to tune the gauge) and frozen before any held-out data (cases set aside in advance to test it).
  3. Timing: each gauge is computed from the stimulation closest to an awakening, ending at the awakening, and a report counts only if it describes what was happening just before waking. Averages over minutes, or windows that end well before waking, do not count.
  4. Rivals recorded: behavioral responsiveness, arousal, and drug level are measured alongside, so the test can say which variable reports actually follow.

Held-out awakenings, not the answer key

PCI's cutoff was set by calibrating it on people whose state was already known from their reports. The conscious side of that calibration included waking, REM dreaming, and ketamine with vivid reports, and the unconscious side included deep non-REM sleep and anesthesia with no reports. Those cases cannot confirm Test A. They are the answer key, not the exam. The test counts only new awakenings that played no part in setting the cutoff, from states named in advance. They are dream reports from non-REM sleep (the sleep stages outside REM), deep sedation with intermittent awakening, sleepwalking and other automatisms, complex seizures, psychedelic states, and covert awareness in unresponsive patients, which brain scans or EEG find in about a quarter of those tested (Bodien et al. 2024). This follows the model of COGITATE, an "adversarial collaboration" in which two rival theories of consciousness (integrated information theory and global workspace theory) wrote down their predictions before the data were collected. Its results (Cogitate Consortium, 2025) challenged key predictions of both.

Holos Prediction

Where the two come apart, reports follow integration. People whose brains are highly integrated but who do not respond are experiencing something. People who respond but whose brains show low integration are not.

How Holos loses

A broken thermometer does not prove heat is fake, so the test separates the gauge from the theory. If one gauge tracks reports and the other does not, the failing gauge is discarded, not Holos. Holos loses if, across independent studies, detailed reports reliably come from periods when both gauges sat clearly below their frozen cutoffs, past any twilight. A report is evidence that something was experienced, and it cannot be explained away as a failure of memory. Nor can it be explained away by a gauge proposed afterward. A new gauge can be tested on new data, but it cannot rescue results already in. The reverse finding, high-integration states that yield no reports, counts against Holos only once report failure can be ruled out (see below).

The confound, and which reports count

Reports require memory, and the states this test targets are precisely those where memory is least reliable. A report of nothing is therefore ambiguous between no experience occurred and experience occurred and was not encoded. The evidence delivers unremembered. The prediction needs unexperienced. This confound is not currently controlled, so silence carries little weight in either direction. It cannot confirm absent experience in low-integration states, and it cannot refute present experience in high-integration ones.

Positive reports escape the memory problem, but not a timing problem. A dream can form in the seconds of waking up and be reported as if it came from deep sleep. That is why the protocol measures right up to the awakening and counts only reports of what was happening just before it. With that control, positive reports carry the test both ways. Where integration is high and behavior is absent, reports of rich experience confirm the prediction without leaning on silence. Where integration is clearly low, a positive report is the losing case. Holos rests Test A on positive reports, and treats silence as suggestive pending a design that calibrates report failure against states with known encoding.

Where the evidence stands

The losing case is not hypothetical. Awakenings from non-REM sleep often produce dream reports, yet non-REM sleep is where whole-brain measures such as PCI fall well below the waking range. Deep sedation is starker. In one study, 82 percent of interpretable awakenings from deep sedation with propofol (a common anesthetic) produced reports of experience (Bajwa et al. 2025), and an earlier study found experiences, mostly dreams, in 84 percent of interviews from anesthetic unresponsiveness (Radek et al. 2018).

The nearest thing to Test A yet run did not favor it. In that sedation study, two complexity measures, one of them a newer variant of PCI (PCIst, on a different scale from the 0.31 cutoff), fell from waking to sedation but did not differ between awakenings with and without experience. A sleep study found the same for a complexity measure of spontaneous EEG (brain activity recorded without stimulation) within light non-REM sleep (Aamodt et al. 2022). Neither settles the question. The sedation study measured over several minutes ending a minute before each awakening, exactly the timing problem above, and it had only five reports of no experience.

Evidence with tighter timing points the other way. When the brain's response to stimulation was measured just before awakening from non-REM sleep, it looked more like the unconscious pattern when subjects then reported nothing, and the more it did, the shorter the dream reports (Nieminen et al. 2016). And work on the neural correlates of dreaming finds that whether a dream is reported, in REM or non-REM sleep, tracks local activity in posterior cortex rather than the state of the whole brain.

That is why the second gauge is local. The maximality condition (an observer forms where integration peaks) allows an aperture smaller than the whole cortex, so an aperture confined to posterior cortex could be integrated above threshold while the whole brain is not. Holos adopts that reading now, after seeing these results, so they cannot count in its favor. Only new data, collected under the protocol above, can confirm or defeat it. A gauge chosen afterward to rescue the prediction would turn the test into decoration.

The data to run it are starting to exist. A shared database released in 2025 pools sleep EEG and dream reports from 505 people and 2,643 awakenings (Wong et al. 2025). It lacks the brain stimulation PCI needs, so it can serve the calibration step, culling candidate gauges, but not the held-out test itself.

What this can and cannot show: this tests a necessary structural condition, not presence itself. It cannot prove an integrated system is an observer, only whether integration is what experience depends on. Holos shares this prediction with other integration-based accounts of consciousness. It is a test Holos could fail, not a signature unique to Holos. Its results also help locate the threshold, under a strict rule. Data used to set a cutoff may tune the measures and weed out weak ones, but never count as evidence. Only data set aside in advance can confirm or refute the claim (see A path to the threshold).

Test B: The twilight narrows with size18

This test is for the transition hypothesis (claim 3 of the threshold), not the core. Genuine transitions in physics look blurred in a small sample and sharpen in a large one. A magnet loses its magnetism at a set temperature (the Curie point), and the larger the magnet, the more abrupt the change. If crossing the threshold is a genuine transition, sudden or continuous, its twilight should likewise narrow as a system grows. Networks of living neurons grown in a dish and connected to simple environments are the cheapest place to look. The DishBrain platform, in which cultured neurons were wired to a simple game of Pong (Kagan et al. 2022), shows the kind of setup, not evidence of experience. Integration is the result, not the knob. Turn a condition such as a drug dose or the network's connectivity, and track an integration measure fixed in advance. Small cultures should cross gradually, and larger cultures, grown the same way, more steeply, as connectivity, feedback, and coupling to their environment increase.

1/4Grown networks

Networks of living neurons grown in a dish, wired to simple environments, are the cheapest place to look. The setup shows nothing about experience by itself.

Why a transition alone is not enough

A sharp transition at loss of consciousness is predicted by ordinary physicalist models too (sudden tipping points and network-wide switch-ons of the kind these systems produce anyway), so observing one confirms nothing specific to Holos. And a cultured network almost always shows some abrupt change in behavior (neurons falling into step and cascades of activity called neuronal avalanches are routine dish behavior), so an experiment that counts any such transition as success cannot fail, and an experiment that cannot fail proves nothing when it passes. What counts is how the steepness scales with size.

How Holos loses

The integration measure and its cutoff are fixed in advance. Holos loses claim 3, though not its core, if the steepness does not grow with size. A steepening that tracks something other than integration, such as arousal, raw activity, or metabolic rate, does not count in its favor. Without a pre-committed measure and a real failure condition, a transition "somewhere" is not evidence. It is decoration.

What this can and cannot show: such transitions would not show that a dish is conscious. They test the shape of the threshold, not presence itself. Sharp drops in integration under anesthesia are useful in the same way, but only as data for Test A, under its rules.

Check C: Observer-relative facts18

This check lives in quantum foundations (the study of what quantum theory says about the world), not in the theory of mind, and its status comes first. It is a consistency check, not a test that could single Holos out. The outcome it anticipates is the one textbook quantum mechanics already predicts. What experiments in this family probe is the family of interpretations Holos belongs to (branching, with no observed event that is a fact for everyone), not Holos alone. The framework's distinctive content, which structures are lived, is a claim about what exists, not something an experiment can show.

1/4The friend

A friend sealed in a lab measures a particle and sees a result.

Extended Wigner's-friend experiments, in which one observer measures another observer who has already made a measurement, already pursue exactly this question. Picture a friend sealed in a lab who measures a particle and sees a result. Outside, Wigner treats the whole lab, friend included, as one quantum system and runs a measurement that probes the friend's result. The 2020 Local Friendliness no-go theorem (an impossibility proof, Bong et al.) shows that three reasonable assumptions cannot all be true together, if the friend's observation counts as a genuine fact. The three are that the friend's result is a fact for everyone (the absoluteness of observed events), that no influence travels faster than light (locality), and that the experimenters' choices are free (freedom of choice). "Local Friendliness" is the name for the three together. Experiments with light, with photons standing in for the friend, break the limit the three assumptions set, just as quantum theory predicts. Holos gives up the absoluteness of observed events. Registered facts are observer-relative, while structural facts and consistency remain intact. In this it sides with branching approaches and borrows one insight from relational ones. Registered facts are indexed to the systems that register them. It does not adopt Relational Quantum Mechanics itself, which rejects the universal state (a single quantum state for the whole universe) that branching requires.

One caveat follows from Holos's own threshold. The "friends" in current experiments with light are far below Φc\Phi_c and register nothing, so these experiments constrain the logical structure of observed events, not registration itself. Holos predicts that repeating them with genuine observers would change nothing physical, a prediction formalized as the standing bet below.

The field is moving toward genuine friends. Wiseman, Cavalcanti, and Rieffel (2023) proposed running a human-level artificial intelligence on a quantum computer as the friend. A September 2026 preprint, not yet peer reviewed, placed agent-like observers on IBM quantum hardware and found the violations intact. Agents that store results and predict their own measurements are still far below Φc\Phi_c by Holos's standard, so the bet remains untested, but this is the road that reaches it.

Holos Expectation

Local Friendliness inequalities keep being violated exactly as quantum mechanics predicts, at every scale the experiments reach, including as the friend is made larger and more complex.

How Holos loses

No experiment in this family can restore the absoluteness of observed events on its own. The theorem is a proof, and a violation only forces a choice among its assumptions. What could go wrong for Holos is dependence on scale. If the violations shrink or vanish as the friend grows toward a genuine observer, beyond what decoherence (the ordinary fading of quantum effects through contact with the surroundings) accounts for, observation is doing something physical, and the standing bet below is lost. A second way to lose lies within a branch. Observers who compare records must find them matching, and a confirmed, irreconcilable mismatch between communicating observers would falsify Commitment 3.

What this can and cannot show: checks Commitment 3 (facts are relational but consistent) against quantum mechanics as it is actually observed. Holos shares its expectation with standard quantum mechanics and with other views that give up absolute observed events, such as Relational Quantum Mechanics. A positive result supports the family, not Holos alone. Running the same test on a quantum computer's chips (superconducting qubits), treating different parts as "observers", would not work as a test. By Holos's own threshold, qubit readouts register nothing, and the result would be ordinary quantum behavior that the branching picture Holos adopts already accounts for.

The standing bet: consciousness adds no new physics

In the version of Holos this site defends, observation is dynamically inert, and that doubles as a bet. A conscious observer obeys the same quantum laws as a photon. Put an integrated system in the measuring role in place of a particle, and Holos predicts no deviation whatsoever. Superpositions (a system being in several states at once) fade because the system becomes entangled with its surroundings, a well-understood process called decoherence, never because someone was home. This does not make experience idle. Under Holos, experience is the inside of the physics, so when the physics does everything, experience is doing its share, not nothing.

1/6Same laws

A conscious observer obeys the same quantum laws as a photon. Put a particle in the measuring role, and the superposition fades.

How Holos loses: unitary quantum mechanics is quantum mechanics in which the state always evolves smoothly and never collapses. If any experiment ever finds a consciousness-linked deviation from it (a superposition that degrades when an integrated observer registers it, beyond what ordinary decoherence accounts for), Holos without collapse is falsified. Observation would make a physical difference after all. The version defended here also loses if experiments confirm collapse of any other kind, such as the size-based collapse that objective-collapse models propose (large objects settling into one outcome on their own, observed or not). That would leave the standing bet about consciousness intact but rule out branching, leaving Holos with collapse. So far every such search has come up empty, and the simplest version has been ruled out.

Two versions, declared now. Holos comes in two versions. They share one core, Axioms 1, 3, 4, and 5, and differ only on quantum physics. Holos without collapse is the version this site defends and the one the bet is about. In it, Axiom 2 is read as unitary evolution. The quantum state always evolves smoothly, every outcome occurs in its own branch, an observer's odds are odds of finding itself in one branch rather than another (self-locating odds), and observation changes nothing it registers. Holos with collapse is the version a lost bet would leave, declared now. Experience would still be the inside of physical activity (Axiom 4), adding no force beyond the physics. That activity would simply include a collapse law. Omega (Axiom 5) would be the universe with its single history, and lived, lit, and unlit would apply within that one history.

Which collapse nature shows decides what the threshold becomes. If a superposed system above the threshold collapses while an equally large one below it does not, collapse tracks integration crossing Φc\Phi_c, and observers matter physically, exactly where Holos says. The threshold becomes the collapse point, detectable for the first time, and registration no longer leaves what it registers unchanged. That is the reading of Holos's named rival, Chalmers and McQueen (2022), who propose that integrated consciousness does collapse the wave function (the quantum state, which collapse would reduce to a single outcome), combining integrated information theory (IIT) with a physical collapse law, and note that versions of the idea could be tested on quantum computers. Holos bets the opposite, so the same experiment would decide between them, and a win for their reading would hand the threshold the strongest confirmation it could get. If collapse tracks size alone, as objective-collapse models propose (large objects collapse on their own, observed or not), the threshold stays dynamically inert and gains nothing.

Declaring the second version now is what separates it from a rescue, but it does not make the two outcomes equal. A lost bet would retire the version defended here, and Holos with collapse would have to earn its own support. Neither version is unfalsifiable. Both share the core, and the core loses through Test A, if experience turns out to track behavior rather than integration. No reading of quantum physics escapes that.

Meanwhile quantum mechanics keeps holding as systems grow. Clusters of more than 7,000 sodium atoms now show quantum interference (Pedalino et al. 2026), and the simplest, parameter-free version of the Diósi-Penrose idea, that gravity collapses superpositions, has been ruled out by a search for the faint radiation it predicts, run deep under the Gran Sasso mountain in Italy to shield it from cosmic rays (Donadi et al. 2021). Versions with an adjustable parameter survive, and Penrose's own version may not predict that radiation.

Some observer-centered frameworks quietly hope consciousness does something physical. Holos formally bets that it does not, and states in advance what losing would cost. The bet is untested so far. A century of placing ever-larger systems into superposition has found no deviation of any kind, but by Holos's own threshold none of those systems was an observer. Photons, molecules, and superconducting circuits all sit far below Φc\Phi_c. That record shows quantum mechanics holding at those scales. It does not yet reach the case the bet is about. The bet is also the one standard physics makes. What makes it worth stating is that Holos, unlike views that need consciousness to act, does not hedge on it. The first experiment to put a system that clearly meets the observer requirements in the friend's role would put it to a real test, provided the friend can be shown, by agreed measures fixed in advance, to be well past the threshold. The consequences are already on the page.

Speculation19

What follows are not predictions. They’re “what if” designs that could emerge if the Holos framework is correct. Each is held to one standard. It must be physically possible and not extremely unlikely, and each says why it is plausible.

We know the familiar hard constraints, which are finite signal speed, noise, and thermodynamics. Across a star system, messages take hours. Across many systems, they take years. A civilization spread thin and bright struggles to act as one, so staying coherent (coordinated as a single whole) rewards compactness, keeping things close, and stability over long spans of time.

1/4Hours

Across a star system, messages take hours.

The Holosian Scale

The Kardashev Scale ranks civilizations by energy use. The Holosian Scale ranks civilizations by integration in a social sense, meaning how well a civilization coordinates as one whole. (This is not the Φ\Phi of the tests above.) The stages below are a map of what “advancement” looks like if coherence, not throughput, is the main objective.

H0: Fragmented

High capability, low coordination. Internal conflict, waste, and short-horizon incentives dominate. Visibility is high because broadcasting is cheap and unmanaged.

H1: Planetary Integration

The civilization becomes coherent at the scale of a world. It can coordinate, self-correct, and sustain long-term projects without collapsing into factional drift.

H2: System Coherence

Coherence survives light-lag (the hours messages take to cross a star system). The civilization functions as one system whose parts need not act in step and shifts from constant broadcast to rare, directed signaling.

H3: Post-Expansion

Physical sprawl stops being the default. Exploration becomes informational first, physical only when inference fails. The outward footprint shrinks even as capability grows.

H4: Deep Integration

The civilization operates like a single high-coherence system with minimal waste and minimal leakage. External visibility fades. What remains detectable is gravitational and thermal, the waste heat no optimization can eliminate.

H5: The Limit

Maximal coherence at minimal waste. This is a limit concept, not a stage any civilization reaches.

Why it is plausible: every stage follows from two certainties, light-speed delay and waste heat, plus one assumption, that coordination pays. The assumption is the weak link. A civilization that never learns to coordinate stays at H0, loud until it ends.

1/4Energy

The Kardashev Scale ranks civilizations by energy use.

Technology

Mesostructures

These are design sketches, not predictions. They are imaginative illustrations of what engineering might look like if the Integration Hypothesis (the companion idea that mature civilizations stay near home, grow efficient, and go quiet in light) holds. Mesostructures, a coinage here, are engineered works far larger than any building but far smaller than the star-enclosing megastructures of science fiction.

The structures below are H3–H4 design patterns, compact enough to stay coherent under light-lag and thermodynamics, and consequential enough to matter without bright sprawl. They span energy generation, active coherence and computation, and long-term continuity.

Holocore

A compact power source, held together by its own gravity, that supplies huge amounts of energy for a very long time and sheds its waste heat in a controlled way.

The likeliest energy backbone for a civilization that stays home is its home star, harvested by nearby collectors (see Aliens). The Holocore is the step beyond. It concentrates energy density rather than surface area, converting mass into controlled output through fusion, regulated accretion (matter fed steadily into a black hole), or extraction of a black hole's spin.

Purpose
  • Provide sustained energy for deep-time computation and preservation
  • Power large-scale modeling, shielding, and entropy management systems
  • Support compact civilizational infrastructure without outward expansion
  • Maintain stability across millennia with minimal maintenance overhead
Design Characteristics
  • Extreme energy density per unit volume
  • Controlled accretion or fusion feed systems
  • Waste heat shaped, delayed, and diluted, but never eliminated. The total thermal output is set by physics, not engineering
  • Compact, so it cannot hide its heat. At modest power it glows faintly in the infrared. At high power it glows brightly, and it is only as quiet as its output allows

The Holocore is infrastructure, not spectacle. If H4 integration suppresses bright sprawl, the energy backbone must be dense and long-lived, and it can be only as quiet as its power allows. Compact and powerful means hot, the same physics that keeps mature systems warm.

Why it is plausible: every step is known physics. Only the engineering is unknown. Fusion releases about 0.7 percent of a mass as energy. Matter falling into a rapidly spinning black hole can release roughly 30 to 42 percent, and the Penrose process can draw out a spinning black hole's rotational energy, up to 29 percent of its mass. A civilization that prizes compactness would prize the densest energy source physics allows.

1/4Fusion

Fusion releases about 0.7 percent of a mass as energy.

Computronium Kernel

A computational core as compact as its cooling allows, built from computronium (matter arranged so that nearly every particle does useful computation) and optimized for coherent, long-horizon modeling rather than raw throughput.

This is not a data center. It is the civilization’s thinking heart, where a unified world-model (a working picture of how everything fits and behaves) is maintained across centuries to millennia.

Purpose
  • Maintaining a single, stable world-model across long horizons
  • Long-range planning (stellar evolution, climate, existential risk)
  • Checking decisions and keeping values and goals from drifting over time
  • Cross-generational model consistency

Note: The Kernel may present as a Dark Node, ordinary matter that has stopped shining (nothing to do with cosmological dark matter), if coherence optimization suppresses the light it gives off. Node describes appearance, not purpose.

Why it is plausible: two known pressures meet here. Light delay rewards compactness. A signal crosses a meter in about three nanoseconds, so smaller thinks faster. Heat punishes it. Power packed too densely cannot be cooled. The Kernel sits where the two balance, as today's chips already do. It need not hold most of a civilization's computing. Erasing information costs less in the cold, so work that need not be fast may run far out, in a cold halo (see The Teeming Dark). The Kernel is the part that must think fast.

1/4Smaller is faster

Light delay rewards compactness. A signal crosses a meter in about three nanoseconds, so smaller thinks faster.

Chrono Vault

A time-optimized preservation structure designed to store civilizational identity, not merely information.

1/4A continuity anchor

Not a library or a backup. It stores values, decisions and their reasons, a way to reboot culture, and who a civilization is.

Not a library or a backup, but a continuity anchor: “If we wake up in 100,000 years, how do we know who we are?”

Purpose
  • Preserving value systems and canonical constraints
  • Storing decision histories and their justifications
  • Rebooting culture after dormancy, collapse, or fragmentation
  • Anchoring identity against drift across deep time

Distinct from the Kernel: the Kernel thinks (active coherence). The Chrono Vault remembers (passive persistence).

Note: Unlike the Kernel, a Vault that stops computing can go cold. It would then not be a Dark Node, which still exports waste heat, but the sleeping case, compact, dark, and close to undetectable.

Why it is plausible: durable storage is ordinary engineering. What is speculative is the motive to pause. The aestivation hypothesis argues computing is cheaper in the colder far future, so a civilization might sleep until then. A published reply argues heat can be dumped cheaply today, so waiting is not required. Collapse, fragmentation, and dormancy remain reasons enough to keep a way back.

Communication

Under known physics, there is no scalable form of real-time interstellar dialogue. Communication converges toward sending large, self-contained packages of information at light speed in tightly focused laser beams.

1/4Packages, not dialogue

No real-time conversation scales between stars. Communication converges on large, self-contained packages sent at light speed in tightly focused laser beams.

At these distances, collaboration is necessarily asynchronous. Civilizations may contribute to shared problem spaces by exchanging durable models, partial solutions, and validated results that remain meaningful even when received centuries or millennia after they were sent. Progress does not depend on shared present time.

Phase-Coherent Beam Transmission

Communication occurs via long-running laser links whose light waves stay in step (phase-coherent), carrying compressed packages that explain their own format, between known or inferred destinations.

  • Purpose: transfer interpretable physical, predictive, and explanatory models across interstellar or intergalactic distances, from small updates to entire civilizational knowledge bases.
  • How it works: laser beams focused as tightly as physics allows, received over long stretches of time, with heavy built-in redundancy so errors can be corrected, and with codes keyed to things every civilization shares, such as physical constants.
  • Payload: layered encodings beginning with mathematics and physical constants, followed by reference frames, compression schemes, and predictive models sufficient to interpret all subsequent data.
  • Why it dominates: photons provide maximum speed, minimal latency, and arbitrarily large total information transfer given sufficient energy and time.
  • Visibility: unless the receiver is aligned in space, time, and frequency, the transmission is effectively invisible. Optical SETI, the search for laser flashes from other civilizations, looks for exactly this kind of signal, one of several technosignatures (detectable signs of technology) astronomers now search for.

Why it is plausible: a visible or near-infrared laser sent through a ten-meter mirror spreads, across ten light-years, to a spot smaller than Earth's orbit. A beam therefore delivers far more signal per watt than broadcasting in every direction, and it is dark to everyone outside its narrow cone.

Exploration

At cosmic scales, most structure is mapped remotely and shared through long-horizon communication. Physical exploration is therefore rare, deliberate, and reserved for regimes where inference alone breaks down.

1/4Mapped from afar

At cosmic scales, most structure is mapped remotely and shared through long-horizon communication.

When physical probes are deployed, they are not explorers in the human sense. They are precision instruments, compact, autonomous, and built to operate alone for decades or longer.

Sentinel Probes

Highly compact, self-contained probes designed to persist in complex environments while gathering high-value physical measurements that cannot be resolved remotely.

Purpose
  • Resolve observational ambiguities by direct measurement where models diverge.
  • Study places too complex or fast-changing to predict from afar.
  • Test and refine predictive models used at civilizational scale.
Technological characteristics
  • Fully autonomous operation, with no expectation of real-time command or intervention.
  • Onboard computation sufficient to evaluate, prioritize, and compress observations on the spot.
  • Preference for passive sensing and indirect interaction over active probing.
  • Extreme energy efficiency enabling long dwell times with minimal thermal or electromagnetic signature.
Operational behavior
  • Long dormant stretches broken by brief, targeted activity.
  • No requirement for interaction with local systems or intelligences.
  • Communication limited to rare, information-dense transmissions rather than a continuous stream of status reports.

Past H3, exploration scales through patience. Sentinel probes exist to watch, not to arrive.

Why it is plausible: Ronald Bracewell proposed exactly this in 1960. A probe parked in a target system studies it in detail no beam can match, and can wait indefinitely. Our own spacecraft already run autonomously for decades.

Gravitational-Lens Observatories

Observation systems that use natural gravitational lenses (a star's gravity bending light like a giant lens) to see in extreme detail without large, glowing infrastructure.

  • Purpose: deep inspection of distant systems already identified as anomalous, interesting, or poorly constrained by existing models.
  • How it works: a star's gravity brings distant light to a focus far away. Instruments placed along that focal line collect light for a long time, trading patience for sharpness.
  • Implication: exploration shifts from surveying everything to interrogating specific questions the shared map cannot yet answer.

Why it is plausible: general relativity places the Sun's focal line beyond about 550 times the Earth-Sun distance, and NASA-funded studies have already designed a mission to image an exoplanet from there.