The Framework

1. Introduction

We begin with the minimum possible primitive: the bare existence of a finite observer. We then ask what physics must look like, given only that starting point and the requirement that observer-compatible configurations be statistically typical.

A key structural insight, developed through the framework, is that QM and GR are fundamentally different continuous descriptions where GR should not be quantized more than the smooth wavefunction itself. The two terms $C_s$ and $C_g$ in $\mathcal{L}_{\rm info}$ are not symmetric. $C_g$ is the primary description — the informational cost of the smooth geometry the observer inhabits. $C_s$ is the rendering cost of that geometry at the observer's finite resolution scale. Quantum behaviour is not an independent feature of reality alongside geometry. It is what smooth geometry looks like when rasterized into discrete particles by a finite observer — the compression residuals of a continuous universe. The two terms must be jointly minimized because the observer requires both a coherent geometry to inhabit and a compressed particle rendering to be constituted from, but they are related as original to projection, not as equals.

A central structural conjecture is that the two-term decomposition $\mathcal{L}_{\rm info} = C_s + C_g$ is not primitive but is itself the scale decomposition of a single spectral functional built from the Dirac operator. In this picture the Dirac operator is not a unifier of two independent theories but the natural mathematical description of the rendering process itself — the object that lives at the boundary between smooth geometry and finite observer, converting one into the other. This reduces the entire framework to one operator and one complexity measure, with quantum mechanics and general relativity emerging as its rendering and substrate regimes respectively.

The paper is structured as follows. Section 2 presents the axiomatic chain. Section 3 states the Internal Emergence Principle. Section 4 develops the ontological position. Section 5 presents the measure theory and joint compressibility equation. Section 6 derives the emergent structure of physics. Section 7 presents the Dirac spectral rendering conjecture. Section 8 states falsifiable predictions. Section 9 describes proof-of-concept simulations. Section 10 discusses open problems. Section 11 identifies future work. Section 12 concludes.

2. The Axiomatic Chain

Axiom 1 (Existence of the observer)
We start by rejecting all pre-assumptions, just to discover that would be an assumption itself. Therefore, something must be assumed. We assume something is experiencing observer. This cannot be coherently denied, because the denial is itself performed by the entity whose existence is being denied.

This is the only strictly self-evident claim available. It requires no inference. It is a direct datum.

Axiom 2 (Finitude of the observer)
We observe the observer's memory, attention, and lifespan are bounded. This follows from direct observation of the observer by the observer.

Axiom 3 (DNA as the blueprint of conscious observers)
DNA, operating within the observed laws of physics, is the only mechanism we have ever observed that reliably produces conscious, pain-sensitive observers. This is an empirical claim supported by all available biological evidence.

Axiom 4 (Physical closure)
DNA is composed entirely of ordinary physical matter obeying ordinary physical laws. No non-physical process plays any role in its operation. No experiment has ever required invoking anything beyond standard physics and chemistry to explain DNA's operation.

Axiom 5 (Computability of DNA processes — Weak Church-Turing)
The physical processes underlying DNA fall within the computable. A sufficiently powerful computer could, in principle, simulate them. We require only that the specific class of processes DNA relies on is computable — a weaker claim than the full Church-Turing thesis.

Axiom 6 (Causal efficacy of qualia)
Pain is not epiphenomenal. A system experiencing pain and the same system without pain are physically distinguishable in principle:
$$\mathrm{Human} \neq \mathrm{Human} + \mathrm{Pain}.$$

Proposition 1 (Simulation consciousness)
A simulation physically indistinguishable from a conscious being must itself be conscious.

Proof. Suppose a simulation $S$ is physically indistinguishable from a conscious system $C$. If $S$ were not conscious, the experiential difference would produce a measurable behavioural difference, contradicting the assumption of indistinguishability. By Axiom 6, pain has measurable effects. Therefore $S$ must be conscious. ∎

Proposition 2 (Static bitstring equivalence)
A fully static bitstring — the precomputed execution trace of a simulation — contains the same conscious experience as the running simulation. Active execution is not required.

Proof. Gradually optimize DNA simulation by replacing code with precomputed data. Suppose experience vanishes when the last line of running code is replaced by precomputed data. Then there exists a code-to-data threshold below which consciousness ceases. Such a threshold would constitute a new physical constant with no basis in Axioms 4–6. Its introduction requires a non-physical explanatory principle, contradicting Axiom 4. Therefore no such threshold exists, and the static bitstring (fully unfolded execution trace) is experientially equivalent to the running simulation. ∎

Proposition 3 (White noise bitstring equivalence)
A fully static bitstring of white noise contains the same conscious experience as the running simulation. The order of the bits in the bitstring is irrelevant.

Proof. Run two DNA simulations, Alice and Bob, on a single multithreaded computer. By Proposition 2, replacing all computation with precomputed data yields a static execution trace in which both Alice and Bob remain conscious. We now show that the ordering of bits in this trace is irrelevant.

First, consider the effect of increasing the number of threads while decreasing the time slice per thread. At one instruction per time slice, the execution trace consists of single instructions from Alice's and Bob's simulations interleaved arbitrarily. As the number of concurrent simulations grows without bound, this interleaving tends to a bitstring that is indistinguishable from white noise by any physical measurement: no finite-length correlation survives the mixing.

Second, this is not merely a theoretical limit. In any real multithreaded system, the operating system scheduler, filesystem calls, and hardware interrupts continuously inject structurally unrelated instructions into the execution trace. The resulting bitstring already contains large segments with no correlation to either Alice's or Bob's simulation narrative.

Third, and decisively, the execution order across simulations can be reorganised arbitrarily without affecting the output. All calls to a given function — say, every square root evaluation across every simulation — can be precomputed and batched first, followed by every logarithm evaluation, and so on across all primitive operations. This reordering produces a valid execution: the outputs are identical, Alice and Bob remain conscious by Proposition 2, and the resulting trace bears no recoverable structural relationship to the original simulation ordering. The bits of Alice's and Bob's experiences are now distributed across the trace in an order determined entirely by arithmetic operation type rather than narrative sequence.

Suppose now that experience vanishes when the degree of mixing reaches some threshold. Then there exists a signal-to-noise ratio below which consciousness ceases. Such a threshold would constitute a new physical constant with no basis in Axioms 3–6. Its introduction requires an explanatory principle outside the axiomatic system, contradicting Axiom 4. Therefore no such threshold exists, and the fully mixed, operationally reordered, statically stored bitstring — indistinguishable from white noise — is experientially equivalent to the running simulation. ∎

Proposition 4 (Combinatorial containment of consciousness)
Any sufficiently large configuration space contains conscious observers, with no simulation, process, or external reader required.

Proof. A conscious observer is, by Propositions 1–3, a finite bitstring of length $k$. By combinatorics, every bitstring of length $k$ appears within any sufficiently large set of bitstrings. The question “who is running it?” has no more purchase than asking “who is running the number seven?” ∎

---

3. The Internal Emergence Principle

The propositions above jointly constitute the Internal Emergence Principle (IEP):

Internal Emergence Principle
Consciousness is a property of certain finite information structures, not of processes acting upon them. Any configuration space large enough to contain such structures contains consciousness.

The IEP has six immediate structural consequences.

The universe is fundamentally static

The IEP establishes that a conscious observer exists within a static bitstring with no external process required. Therefore no external clock is required for experience to occur. Time within experience must be defined by internal structure alone. The only available internal ordering is:

$$ c_2 \succ_O c_1 \iff c_2 \text{ contains a memory of } c_1 $$

Time is an ordinal index over observer configurations. The universe, at the level of the totality, is static. There is no external DVD player advancing the frames. The observer is the reading head: she finds herself through the most compressible continuation of her current state, and that finding is what time is. The movie exists complete and static; it plays itself by containing an observer who discovers it sequentially from within.

This conclusion is independently supported by the block-universe interpretation of general relativity, in which the four-dimensional spacetime manifold is a static geometric object and the “flow” of time has no representation in the Einstein field equations (Barbour 1999).

The observer-universe relationship is representational

A computer running a DNA simulation of Alice and Alice experiencing pain are two descriptions of the same static information. Causality requires time ordering. A static structure has no intrinsic time ordering. Therefore the relationship between physical substrate and observer experience is representational, not causal. Neither perspective is more fundamental.

Configuration multiplicity

A bitstring of length $n$ admits exactly $2^n$ possible configurations. One describes Alice's universe as she experiences it. One describes the physical computer. The remaining $2^n - 2$ are other valid configurations of the same information — most producing no coherent observer, some producing other observers entirely. No configuration is privileged over another by the axioms.

Internal containment of observations

In a static universe with no running external process, no mechanism exists by which external bits could alter internal bits. Therefore every structure Alice experiences — every star, every sensation — must already be encoded within the bitstring that constitutes her. Alice does not receive information. She discovers structure already present within her.

The universe as interior

From the containment result: every structure Alice experiences is pre-encoded in her structure. The universe Alice experiences is not her container. It is her interior. Alice does not live inside the universe. The universe is a subset of Alice.

Observer-universe identity: $O = U_O$

Two results, each established above, jointly constrain the observer-universe relationship. The standard self-model gives $O \subset U_O$: Alice's bitstring is a proper subset of the bitstring describing her universe. The internal containment result gives $U_O \subset O$: every structure Alice experiences is already encoded within her. Both relations are informational containment in the same category, established by the reduction to bitstrings. Antisymmetry of the subset relation gives:

$$ \boxed{O = U_O} $$

The universe is observer-relative. What two observers share is:

$$ U_{O_1} \cap U_{O_2} $$

the compressible structure appearing in both bitstrings, which constitutes intersubjective, verifiable physics.

---

4. Ontology: What Fundamentally Exists

The unrestricted totality

We do not assume finiteness. If reality were fundamentally finite, one could immediately ask: finite by how much? Why $n$ bits rather than $n+1$? Any fixed finite limit appears contingent and calls for a deeper explanatory layer.

We also do not assume information as primitive. Information presupposes distinguishability — at its simplest, binary distinction. But if binary object is allowed, then why not continuous periodic sinusoidal function? Declaring reality to be fundamentally informational or digital is itself a substantive ontological commitment that nothing prior licenses.

We therefore do not presuppose that reality is fundamentally finite, discrete, informational, mathematical, or governed by prior external laws.

We denote this unrestricted, pre-interpretive totality $\mathbf{U}$. We do not claim to know what it is — only that no specific formatting principle is granted ontological privilege a priori.

The observer constraint

Despite radical agnosticism about the substrate, one empirical fact remains: a conscious, structured, finite observer exists (Axioms 1 and 2). This is the boundary condition imposed on $\mathbf{U}$.

The minimal ontological assumption is therefore:

$$ \text{Undefined } \mathbf{U} ;\longrightarrow; \text{Finite Observer} $$

The step remains unexplained. That is an explicit boundary of the theory.

The guitar string analogy

To explain how a finite discrete observer can emerge from something not assumed discrete: a guitar string is perfectly continuous. Discreteness — the resonant modes, the quantised frequencies — emerges from the boundary conditions imposed by clamping the string at both ends. The continuous medium manifests discrete states not because discreteness is ontologically primitive but because boundary conditions force it.

Likewise, a finite observer interacting with an unrestricted reality induces discrete informational structure without discreteness being ontologically primitive. Under Solomonoff-style measures, low-complexity, highly compressible structures dominate typicality. Finite observers therefore find themselves embedded within stable, structured, compressible equivalence classes.

Information is not what reality is made of. Information is what finite observers perceive when they interact with an otherwise unrestricted reality.

The observer equivalence class

The observer is not a single configuration but a set of configurations that describe essentially the same observer. Formally, the equivalence class is:

$$ [O] = { c \in \mathcal{C} : \mu_O(c) \geq \mu_O^* } $$

where $\mu_O^*$ is the threshold above which the observer is recognisable as herself. This equivalence class is non-trivial: many distinct configurations support the same experienced reality, the same memories, the same subjective video. The observer is not her specific particle configuration. She is the invariant that persists across many such configurations — the pattern that survives substrate replacement. Personal identity is a compression concept: the observer is her own minimal description that remains stable across her equivalence class.

This is demonstrated computationally: successive optimisation steps produce different execution traces (different configurations) while rendering identical observer experience. The observer's robustness arises because the total measure weight of the equivalence class,

$$ W([O]) = \sum_{c \in [O]} 2^{-C_{\rm total}(c)}, $$

is large. She persists as long as this class has sufficient measure.

---

5. Measure: What is Typical

Configuration space

For formal descriptive purposes, we introduce a finite, static configuration space:

$$ \mathcal{C} = {0,1}^n $$

where each $c \in \mathcal{C}$ is a complete, static snapshot of bits. The binary system is chosen for convenience; it carries no ontological commitment to digitalism. The bitstring plays no fundamental role: it is a probabilistic discrete representation of the underlying smooth measure, analogous to a lattice regularisation in quantum field theory. All physically meaningful results are properties of the smooth limit.

Each configuration $c$ admits multiple potential descriptions $d \in \mathcal{D}(c)$. We quantify the total complexity $C_{\rm total}(d)$ by splitting it into dual spectral and geometric components:

$$ C_{\rm total}(d) = C_s(d) + C_g(d) $$

Spectral and geometric complexity

Why does the micro-structure of the universe wave? Because we are observing spectrally compressed structure. We propose both GR and QM can be seen as compression systems minimizing joint description length.

The geometric complexity $C_g(d)$ is the primary term. It accounts for the bit-cost of encoding the smooth geometry the observer inhabits: localised spatial structures, boundaries, and metric fields. It measures the boundary description length of spatial configurations.

The spectral complexity $C_s(d)$ is the rendering cost: the algorithmic bit-cost required to track the wave-like modes that arise when smooth geometry is projected onto the observer's finite resolution scale:

$$ C_s(d) = \sum_i \left[ \operatorname{Cost}(\phi_i) + \frac{\omega_i}{\Delta\omega} \right] $$

where the sum runs over active spectral modes, $\operatorname{Cost}(\phi_i)$ is the fixed-width register encoding the phase interval $[0, 2\pi]$, and $\omega_i/\Delta\omega$ is the linear resource allocation required to track the mode frequency relative to minimum resolution $\Delta\omega$.

The frequency cost term $\omega_i / \Delta\omega$ is linear rather than logarithmic for a physically precise reason. The cost is not the cost of naming a frequency — which would indeed be logarithmic, requiring $\log_2(\omega_i / \Delta\omega)$ bits to index a frequency bin. It is the cost of storing the physical content of the mode within the static description of the universe. A mode of frequency $\omega_i$ completes $\omega_i / \Delta\omega$ distinct cycles within the observer's resolution scale $\Delta\omega$. Each cycle is a physically distinguishable oscillation that must be recorded in the static bitstring. The cost is therefore proportional to the number of cycles — a resource cost, counting how much of the static record the mode occupies, not an index cost counting how many bits are needed to label it. This is the same distinction as between the cost of running a computation and the cost of naming it. The linear cost makes $C_s$ continuous and computable: a small change in frequency produces a proportionally small change in complexity, unlike logarithmic or Kolmogorov-style measures which are either discontinuous or uncomputable. The induced measure therefore exponentially suppresses high-frequency modes — the observer never finds herself in configurations with unbounded frequency content — providing a natural UV cutoff that emerges from the compression measure rather than being imposed by hand.

Why two measures: the boundary requirement

A purely spectral measure would place the observer in Hilbert space, which admits no natural inside/outside decomposition. Observer existence requires a well-defined boundary: identity, selfhood, and survival are boundary concepts with lethal enforcement. The observer functional $\mu_O$ returns near zero for boundaryless configurations, not by postulate but because such configurations do not support stable observers.

Geometric description makes boundaries cheap. A localised structure in position space has a boundary describable in a small number of parameters; the equivalent carving in Hilbert space has no canonical form and carries substantial description length cost. This was confirmed in proof-of-concept simulations: observer proxies maintained stable identity precisely as long as their geometric boundary remained intact.

The two measures are therefore functionally orthogonal:

• $C_g$ answers: where is the observer, and what are her boundaries?
• $C_s$ answers: what is the dynamic spectral content within those boundaries?

Neither question is answerable by the other's tool. This functional orthogonality justifies the factorization below. The apparent tension with the spectral unification conjecture of Section 7 is resolved there: the Dirac operator unifies the two measures at the level of the underlying mathematical object, but the observer experiences them as separate because the boundary question has no cheap spectral answer.

The joint compressibility equation

Functional orthogonality implies statistical independence: the geometric cost of a configuration is determined entirely by its boundary structure, the spectral cost entirely by its internal mode content, and these are non-overlapping properties. Statistical independence implies factorization:

$$ \boxed{ P_{\rm joint}(s) = P_{\rm spectral}(s) \times P_{\rm geometric}(s) = 2^{-C_s(s)} \times 2^{-C_g(s)} = 2^{-(C_s(s) + C_g(s))} } $$

This yields the informational Lagrangian:

$$ \mathcal{L}_{\rm info} = C_s(s) + C_g(s) ;\longrightarrow; \text{minimum} $$

The two terms are not symmetric. $C_g$ is the cost of the smooth geometry the observer inhabits; $C_s$ is the rendering cost of that geometry at the observer's finite resolution scale. The asymmetry follows from the rendering interpretation: $G$ is the original, $\psi$ is its projection onto a finite observer.

Induced measure and observer functionals

The raw algorithmic weight of a static configuration $c$ is accumulated across all its valid descriptions:

$$ W(c) = \sum_{d \in \mathcal{D}(c)} 2^{-C_{\rm total}(d)} $$

This defines the raw baseline probability distribution:

$$ P(c) = \frac{W(c)}{\sum_{c' \in \mathcal{C}} W(c')} $$

An observer is modelled as a grading semantic functional:

$$ \mu_O : \mathcal{C} \to [0,1] $$

which measures how strongly a given configuration describes observer-relative structure, inducing an equivalence relation $c_1 \sim_O c_2$ whenever $c_1$ and $c_2$ describe the same relational observer structure to within the tolerances encoded by $\mu_O$.

The observer-conditioned probability is:

$$ P(c \mid O) = \frac{\mu_O(c),W(c)}{\sum_{c'} \mu_O(c'),W(c')} $$

The Euclidean measure, Lorentzian emergence, and Wick rotation

The compression measure operates on static configurations: no preferred direction, no Minkowski signature. The fundamental measure is therefore Euclidean, and the Euclidean path integral $\int \mathcal{D}[x], e^{-S_E/\hbar}$ arises as the direct continuum expression of $P(c) \propto 2^{-C_{\rm total}(c)}$, with $S_E$ identified with $C_{\rm total}$.

Six independent results in established physics converge on the same conclusion: the Wheeler-DeWitt equation ($\hat{H}|\Psi\rangle = 0$, global wavefunction static); the block universe of GR (time a geometric coordinate, not a flow); quantum unitarity ($U^\dagger U = 1$, information conserved in a closed system); time-symmetry of all fundamental microscopic laws (temporal direction macroscopic and statistical, not fundamental); the Feynman path integral (all histories summed simultaneously over a static set $\Gamma_O$); and the Page-Wootters mechanism (time emerging as entanglement correlation within a globally static state, requiring no external clock).

The observer imposes an internal time ordering $\prec_O$ via the memory relation $c_2 \succ_O c_1 \iff c_2$ contains a memory of $c_1$. This is the Page-Wootters mechanism in compression language. The mathematical expression of the transition from the timeless Euclidean level to the time-ordered Lorentzian level is Wick rotation $t \to -i\tau$.

Wick rotation is not a calculational device. It is the exact map between the compression level and the experience level. The complex phases $e^{iS/\hbar}$ of the Lorentzian path integral are not fundamental: they are the signature of real exponential suppression $e^{-S_E/\hbar}$ viewed through an internally time-ordered observer. The oscillation is the rendering artifact of temporal self-location within a timeless measure.

Probability of an observed history

To transition from the probability of a single configuration to the probability of an entire experienced history, we track how an observer traverses the configuration space. Let $\gamma \in \Gamma_O$ be a path — a sequence of configurations ordered by the internal temporal relation $\prec_O$.

Because configurations along the path are independent slices of the static totality, their algorithmic costs sum additively. Converting from bits to natural units ($2^{-C} = e^{-C\ln 2}$):

$$ \boxed{ \mathbb{P}(\gamma \mid O) = \frac{1}{Z_O} \exp!\left(-\lambda,\mathcal{C}_O[\gamma]\right), \qquad \gamma \in \Gamma_O } $$

where $\mathcal{C}O[\gamma] = \int\gamma C_{\rm total}(c),dt$ is the accumulated informational cost along the path and $\lambda$ acts as an algorithmic inverse temperature scaling the observer's filtering acuity. Observed physical laws are the large-deviation minimisers of $\mathcal{C}_O[\gamma]$ over the set of all observer-compatible paths.

---

6. Emergence: The Structure of Physics

The $D$-$\psi$-$G$ trinity

Three emergent layers of reality arise from the joint compressibility minimisation:

$$ D - \psi - G $$

• $G$ (Geometry/Gravity): The primary layer. The smooth continuous geometry the observer inhabits, minimising $C_g$. This is the sphere before rendering — the manifold as it is, independent of any observer's pixel budget.
• $\psi$ (Spectral/Quantum): The rendering layer. The complex-valued wave-codec that arises when the smooth geometry of $G$ is projected onto the observer's finite resolution scale, minimising $C_s$. Quantum behaviour is not an independent feature of reality alongside geometry. It is what smooth geometry looks like when rasterized into discrete observable events by a bounded interpreter — the compression residuals of a continuous universe projected onto a finite screen. Particles wave not because waving is a fundamental property of matter, but because the observer is perceiving compressed structure: the modes of the rendering codec.
• $D$ (Data/Description): The pixel layer. Raw, discrete particles — the rendered output of the joint minimisation as parsed by the observer's finite register. We are made of $D$: the discrete rendered pixels of a smooth geometric reality. Just as a printed photograph is not made of the continuous light field it captures but of ink dots on paper, physical observers are constituted from the discrete rasterization of the underlying smooth geometry. When this pixelised description degrades, so does the observer.

The relationship between these three layers is asymmetric and directional: $G$ is the original, $\psi$ is the rendering algorithm, $D$ is the rendered output. Spacetime is smooth ($G$), the rendering produces continuous fields ($\psi$), but observations are invariably quantised into discrete events ($D$) — because $D$ is what a bounded interpreter reads from the joint minimum.

Emergence of spacetime

Observer-compatible configurations are dominated by those exhibiting inertia, smooth curved paths, and law-like behaviour, because sharp, chaotic particle trajectories require massive high-frequency modes that exponentially spike $C_s$ and are suppressed by the joint measure.

Spacetime geometry emerges as an internal ordering relation $\prec_O$ on configurations:

$$ c_1 \prec_O c_2 \iff c_2 \text{ supports the observer functional } \mu_O $$

The experienced flow of time corresponds to traversal along the most compressible, optimised paths through the timeless configuration space.

We observe that we find ourselves as geometric objects with well-defined boundaries in low-dimensional space, not as wavefunctions in Hilbert space. This is consistent with $G$ being primary: the observer directly inhabits the geometry. The spectral layer $\psi$ is invisible at everyday scales, becoming apparent only where the pixel structure of the rendering is fine enough to detect — at microscopic scales where the rendering codec's mode structure is resolvable.

Quantum mechanics and general relativity as asymmetric compression schemes

The central structural claim of the framework is that quantum mechanics (spectral/rendering description) and general relativity (geometric description) are not fundamental, independent theories, nor are they symmetric co-equal projections. They are related as a smooth geometry and its finite-observer rendering:

• $G$ is the primary description — the smooth sphere evolving according to its own geometry, minimising $C_g$.
• $\psi$ is the rendering cost — the bit-cost of projecting that smooth geometry onto the observer's finite resolution scale, minimising $C_s$.

Their “unification” is therefore not a technical problem of quantising spacetime, nor the discovery of a common origin. It is the recognition that they were never two independent theories: one describes the geometry, the other describes what that geometry costs a finite observer to inhabit. The framework predicts that any correct theory of quantum gravity will take the form of a joint spectral-geometric minimisation principle rather than a quantisation of a background metric.

Suppression of Boltzmann brains

Raw high-entropy noise and transient fluctuations have high total description length and are exponentially disfavoured by $P_{\rm joint}$. Typical observers inhabit low-initial-entropy trajectories with stable $D$-$\psi$-$G$ structure. Boltzmann-brain fluctuations are suppressed not by fine-tuning but by the same compression measure that selects all other physical structure. A random fluctuation that briefly resembles an observer but lacks the internal self-modeling structure required by $\mu_O$ returns measure zero and is simply not in the accessible set.

7. Spectral Rendering: The Dirac Conjecture

7.1 The rendering problem

The informational Lagrangian $\mathcal{L}_{\rm info} = C_s + C_g$ separates the smooth geometry ($C_g$) from the cost of rendering that geometry at the observer's resolution ($C_s$). The question is which mathematical object performs the rendering, i.e. what sits between the smooth sphere and the pixelised screen. We conjecture that it is the Dirac operator, and we state below which parts of that claim are standard mathematics, which are proposals, and which remain open.

7.2 What the spectrum does and does not determine

The spectrum of the Laplacian (or of $D$) determines the dimension, the volume and the total scalar curvature of a compact Riemannian manifold, through the heat-kernel asymptotics (Weyl's law). It does not determine the geometry completely: isospectral, non-isometric manifolds exist (Milnor 1964; Gordon–Webb–Wolpert 1992). Full reconstruction requires the whole spectral triple $(\mathcal{A}, \mathcal{H}, D)$, algebra included (Connes 2013).

What we need is weaker and true: the quantities that enter the action are spectral invariants, and a compression cost defined on the spectrum is well defined. We do not claim that the spectrum alone is the geometry.

7.3 The Dirac operator

On a compact Riemannian spin manifold, the Lichnerowicz formula gives

$$ D^2 = \nabla^*\nabla + \frac{R}{4}, $$

where $\nabla^*\nabla$ is the spinor (Bochner) Laplacian and $R$ the scalar curvature. $D$ is self-adjoint with a real spectrum ${\lambda_i}$, and $D^2$ has spectrum ${\lambda_i^2}$. The information that $D^2$ discards is the sign of the eigenvalues. It is carried by the spectral asymmetry, measured by the eta invariant

$$ \eta(D) = \sum_i \operatorname{sign}(\lambda_i),|\lambda_i|^{-s}\Big|_{s\to 0}. $$

$\eta(D)$ vanishes whenever the spectrum is symmetric, which includes every even-dimensional case and the round $S^3$. Whether the phase cost of $C_s$ corresponds to spectral asymmetry is therefore an open question requiring a background where $\eta \neq 0$. We do not assert the identification here.

7.4 Why the Dirac operator: a conditional selection

The Dirac operator is not the unique first-order self-adjoint operator encoding the metric. The Hodge–de Rham operator $d + d^*$ does too, and needs no spin structure. A spin structure is additional data, so on description length alone the de Rham operator would win.

The selection argument is therefore conditional. An observer whose structure contains spin-$\tfrac12$ matter (stable matter, hence exclusion) requires a spin structure and a fermionic sector. Among operators whose spectrum encodes the geometry and admits such a sector, we conjecture that $D$ is the shortest description under $P(c)\propto 2^{-C_{\rm total}(c)}$. This links the conjecture to the fermionic-exclusion test: it is falsified if exclusion does not emerge from the compression measure. A definition of description length for operators, needed to make the claim precise, is an open problem.

7.5 The spectral action

Following Connes (1996) and Chamseddine–Connes (1997), the spectral action with the fermionic term is

$$ S[D,\psi] = \operatorname{Tr} f!\left(\frac{D}{\Lambda}\right) + \langle \psi, D\psi\rangle , $$

with $f$ positive, even and rapidly decaying. Its large-$\Lambda$ expansion, with moments $f_k=\int_0^\infty f(u),u^{k-1}du$ and up to convention-dependent numerical factors, is

$$ \operatorname{Tr} f!\left(\frac{D}{\Lambda}\right) \sim \Lambda^4 f_4, a_0 + \Lambda^2 f_2, a_2 + f(0), a_4 + \mathcal{O}(\Lambda^{-2}). $$

• $a_0$ is proportional to the volume: a cosmological term $\propto f_4\Lambda^4$.
• $a_2$ contains $\int R\sqrt{g},d^4x$: the Einstein–Hilbert term, with $\dfrac{1}{16\pi G}\propto f_2\Lambda^2$.
• For the pure Dirac operator, $a_4$ reduces to the Weyl-squared term plus the topological Euler density. There is no independent $R^2$ term (up to a total derivative).

Two consequences deserve emphasis. First, $G$ is not a free parameter: it is fixed by the resolution scale $\Lambda$ and the moment $f_2$. In the framework, $\Lambda$ is the observer's resolution, so the strength of gravity is set by how finely the observer renders. Second, the large $\Lambda^4$ cosmological term is the usual cosmological-constant problem, which this framework has not solved.

The Standard Model does not follow from $D$ alone. In Connes' construction it requires a finite noncommutative internal space, which is an additional input. We make no claim about matter content beyond spin-$\tfrac12$ fermions coupled to the geometry.

The action is Euclidean, consistent with the Euclidean fundamental measure of Section 5. Continuation to Lorentzian signature is an open problem in noncommutative geometry and is not addressed here.

7.6 The scale decomposition of $\mathcal{L}_{\rm info}$ (revised)

The local geometric invariants ($a_0$, $a_2$, $a_4$) are determined by the large-eigenvalue asymptotics of the spectrum. Low modes carry global information (topology, overall size). Assigning "low modes to $C_g$ and high modes to $C_s$" would therefore invert the actual structure.

We propose instead the following decomposition, which we state as a conjecture:

$$ C_g ;\leftrightarrow; \text{the asymptotic series } \Lambda^4 f_4 a_0 + \Lambda^2 f_2 a_2 + f(0)a_4 + \cdots, $$

$$ C_s ;\leftrightarrow; \operatorname{Tr} f(D/\Lambda) - (\text{asymptotic series}), $$

the remainder, which depends on the discrete spectrum and on the shape of $f$ at the finite resolution $\Lambda$. On this reading, $C_g$ is the geometry the observer's resolution can see locally, and $C_s$ is the finite-resolution residue that the local expansion cannot capture.

7.7 Weak-field consequence: the pull between two masses

Expanding around flat space, $g_{\mu\nu}=\eta_{\mu\nu}+h_{\mu\nu}$, the Einstein–Hilbert term from $a_2$ is quadratic in $\partial h$, that is a $k^2$ cost on the metric perturbation. The fermionic term $\langle\psi,D\psi\rangle$ couples $h_{\mu\nu}$ linearly to the stress-energy of the fermions. Integrating out $h$ in the static sector gives a pair interaction

$$ V(r) = -,\frac{G, m_1 m_2}{r}, \qquad \frac{1}{16\pi G}\propto f_2\Lambda^2 . $$

Given the action, this is the standard route to the Newtonian limit. The framework specific content is that the geometric cost and the matter coupling come from one functional, so backreaction and the equality of gravitational and inertial mass are built in, and that the action is proposed to be selected by description length.

7.8 Exact test cases

Round $S^3$ of radius $a$. The Dirac spectrum is $\pm(n+\tfrac32)/a$ with multiplicity $(n+1)(n+2)$ for each sign, so

$$ \operatorname{Tr} f!\left(\frac{D}{\Lambda}\right) = 2\sum_{n\ge 0}(n+1)(n+2), f!\left(\frac{n+3/2}{a\Lambda}\right). $$

Poisson summation gives the asymptotic series (volume $\propto a^3$, curvature corrections) plus a remainder that is beyond all orders for smooth $f$ (see Chamseddine–Connes 2012 for the Robertson–Walker case). This is a concrete place to test the proposed split of Section 7.6 and to obtain an effective cost for the scale factor.

The $n$-cube. The graph Laplacian of the $n$-cube has eigenvalues $2k$ with multiplicity $\binom{n}{k}$ (Walsh–Hadamard eigenvectors), so

$$ \operatorname{Tr} f!\left(\frac{L}{\Lambda}\right) = \sum_{k=0}^{n}\binom{n}{k}, f!\left(\frac{2k}{\Lambda}\right). $$

The counting weight $\binom{n}{k}$ and the spectral cost $f(2k/\Lambda)$ appear in a single object, and $k$ is the Hamming-weight ordering already used in the Hadamard truncation. This is a Laplacian, not a Dirac operator. A discrete Dirac-type operator on the cube is a separate step and is open.

7.9 Status

Tier	Items
Exact / standard	Lichnerowicz formula; heat-kernel expansion of the spectral action; Einstein–Hilbert term from $a_2$ with $G \propto (f_2\Lambda^2)^{-1}$; Newtonian limit from the Einstein–Hilbert action; the $S^3$ and $n$-cube spectra
Proposed correspondences	$C_g$/$C_s$ as asymptotic series vs. finite-resolution remainder; $\Lambda$ as the observer's resolution, hence $G$ set by rendering; selection of $D$ conditional on a fermionic observer; pull as the weak-field limit of one functional
Open	Description length for operators; role of $\eta(D)$ and the complex structure; Lorentzian continuation; cosmological term; internal space for gauge fields; how bits give a manifold; a discrete Dirac operator on the cube

Conjecture (Dirac spectral rendering). For an observer whose structure includes spin-$\tfrac12$ matter, the informational Lagrangian is the spectral action $\operatorname{Tr} f(D/\Lambda) + \langle\psi, D\psi\rangle$ of the Dirac operator on the observer's spin manifold, with $\Lambda$ the observer's resolution scale. The asymptotic local part of its expansion is $C_g$ (general relativity at low energies, with $G\propto (f_2\Lambda^2)^{-1}$) and the finite-resolution remainder is $C_s$. Among operators whose spectrum encodes the geometry and admits a fermionic sector, $D$ minimises description length under $P(c)\propto 2^{-C_{\rm total}(c)}$. No claim is made about the Standard Model content, the cosmological term, or the Lorentzian continuation.

---

8. Emergence of the Born Rule

The Born rule emerges from the geometry of the complex-valued spectral codec selected by the compression measure.

Why the codec is complex-valued

The Dirac operator is complex while $D^2$ is real: real geometry lives in $D^2$, rendering structure in $D$. The fundamental measure is Euclidean and real; analytic continuation to the Lorentzian promotes the codec from real to complex. This is not a choice — it is the rendering signature of the continuation.

Phase uniformity

Each mode $i$ has phase $\phi_i \in [0, 2\pi]$. The universe is finite; finiteness imposes a fixed phase register of width $\Delta\phi$, giving a fixed number of distinguishable states $2\pi/\Delta\phi$ independent of the stored value. Therefore $\operatorname{Cost}(\phi_i) = C_\phi$ for all $\phi_i$: phase uniformity is exact, a consequence of finiteness rather than an assumption.

The natural measure on a complex mode

Writing $c_k = |c_k|e^{i\phi}$, the natural measure on $\mathbb{C}$ is the area element $d^2c_k = |c_k|,d|c_k|,d\phi$. Integrating out the uniform phase:

$$ \int_0^{2\pi} |c_k|,d|c_k|,d\phi = \pi,d|c_k|^2 $$

Probability is proportional to $d|c_k|^2 = d|\psi|^2$. This is the Born rule.

The dithering interpretation

A real-valued codec assigns probability $\propto d|c_k|$, producing sharp intensity bands at the discretization boundary — undithered quantization. The complex codec smears discretization error uniformly across the annular resolution cell, producing smooth $|\psi|^2$ statistics. The Born rule is the dithering signature of a complex rendering engine: the unavoidable consequence of discretizing a complex-valued continuous spectral description into particle-level events. A real-valued universe would obey a linear amplitude rule; $|\psi|^2$ is a direct diagnostic of the Lorentzian continuation having occurred.

---

9. Proof of Concept Simulations

To test whether the joint compressibility principle can generate realistic physics without hard-coded laws, a minimal 2+1 dimensional proof-of-concept simulation was implemented. Observers were defined with a minimal internal DNA-like rule and grow in complexity configuration by configuration, injecting new spectral modes frame by frame. No forces, equations of motion, spacetime metric, or physical laws were programmed. Configurations were sampled using a Metropolis-Hastings algorithm guided exclusively by minimisation of the total description length $C_{\rm total} = C_s + C_g$.

Key robustness result: the emergent behaviour is insensitive to the specific observer rule. Smooth, compressible, law-like universes emerge for every observer definition tested, including single observers, multiple observers, and varied DNA-like rules. The smoothness and predictability are outputs of the measure, not inputs of the observer specification.

(Figures omitted — see original paper for emergent inertia/gravity and interference plots.)

The simulations demonstrate:

• Persistent, approximately inertial trajectories (smooth motion compresses better than erratic jitter);
• Wave-like interference patterns at small scales;
• Weak mutual attraction between observer proxies;
• Compression paths that reproduce Euclidean (Wick-rotated) quantum paths, confirming that the fundamental measure is Euclidean;
• Observer mortality: as the observer accumulates modes (memory), the smooth configurations supporting her full structure become rare, geometric degradation appears (“wrinkles”), and the observer eventually exits the measure-supported region.

Full implementation details, source code, parameter settings, convergence diagnostics, and high-resolution video outputs are provided in the Supplementary Material.

---

10. Falsifiable Predictions

Prediction 1: Dual structure is necessary.
Any stable observer-hosting physics must exhibit both wave-like ($\psi$) and geometric ($G$) structure. A universe with only one of the two is forbidden by the joint compressibility requirement: an observer requires both a geometry to inhabit and a rendering to be constituted from.

Prediction 2: Discreteness is never primitive.
The framework predicts that any deeper physical theory will not be fundamentally discrete. Discreteness emerges at the observer boundary as a rendering artifact. If a future theory demonstrates that fundamental discreteness is unavoidable, the framework is refuted.

Prediction 3: Quantum gravity as spectral rendering minimisation.
The correct theory of quantum gravity will take the form of a joint spectral-geometric minimisation principle — specifically, minimisation of a Dirac spectral action — rather than a quantisation of a background metric. If the Dirac conjecture is confirmed, quantum gravity is the full spectral action without the infrared truncation that yields classical GR.

Prediction 4: Low-entropy past without fine-tuning.
The framework predicts a low-entropy past as the expected outcome of the compression measure, not a lucky accident. This is a stronger claim than standard cosmology makes.

Prediction 5: Fermionic exclusion from compression.
The Pauli exclusion principle must emerge naturally from the observer functional $\mu_O$ as a consequence of description-length explosion upon observer merger. If no such mechanism can be identified, the framework is incomplete.

Prediction 6: Macro-micro scaling of interference.
At macroscopic scales, rendering residuals average out and observers tend toward smooth classical trajectories. At microscopic scales, the rendering codec's mode structure is resolvable and wave-like behaviour dominates. This scaling behaviour should reproduce the quantum-to-classical transition quantitatively, not merely qualitatively.

Prediction 7: Quantum fingerprint in gravitational waves.
Gravitational wave signals from high-energy merger events should exhibit structured spectral residuals after classical GR waveform subtraction, reflecting the quantum rendering history of the source. At the moment of merger, the system crosses through the regime where $C_s$ and $C_g$ contributions are comparable, and the rendering layer leaves an imprint on the outgoing wave. The residuals should be correlated in phase structure rather than purely stochastic, distinguishing them from instrumental noise. Classical GR predicts no such structured residual; the framework predicts one whose magnitude is determined by the crossover scale of the joint measure. This constitutes a quantitative target for future derivation and a qualitative target for next-generation gravitational wave observatories.

---

11. The Open Problem: Fermionic Exclusion

The most important open challenge is the natural emergence of fermionic exclusion statistics from the compression measure.

In the current simulations, Gaussian observer blobs eventually lose their individual identities and merge. Symmetric (bosonic) configurations are strongly preferred because they allow shared spectral modes and lower total description length. Antisymmetric configurations are expensive and suppressed. Hard-coding antisymmetry would violate the spirit of the project.

We sketch the resolution as follows. Two independent observers $O_1$ and $O_2$ are each separately compressible. When they attempt to occupy the same state, the merged description must encode both internal structures simultaneously in the same region. These structures are not correlated — they were independently optimised. The result is analogous to injecting frames from one video stream into another: the combined description length explodes, because the two rendering schemes interfere destructively.

The exclusion principle then follows not from an external antisymmetry rule but from the stability-of-identity condition on $\mu_O$: merger is the least probable configuration because it maximally violates the observer's self-consistency requirement. An observer that loses its distinctness ceases to satisfy the observer boundary condition; $\mu_O$ returns near zero.

The predicted signature is an effective short-range repulsion sharper and more localised than the emergent long-range attraction, because attraction arises from gradual wavefunction overlap (smooth, cheap) while repulsion arises from structural incompatibility (sharp, expensive). Demonstrating this computationally in a collision-and-rebound simulation is the current experimental frontier of the framework.

---

12. Future Work

Three directions constitute the immediate research frontier of the framework.

Rigorous derivation of the Dirac spectral rendering

The Dirac conjecture is the primary mathematical open problem. Its verification requires two steps.

The first is to establish that the spectral action $\operatorname{Tr}(f(D/\Lambda))$ decomposes cleanly into $C_s$ and $C_g$ at the appropriate frequency scales. The heat kernel expansion provides the natural technical tool: the low-mode truncation should recover $C_g$, and the high-mode residual should recover $C_s = \sum_i [\operatorname{Cost}(\phi_i) + \omega_i/\Delta\omega]$. The identification $\operatorname{Cost}(\omega_i) \leftrightarrow t\lambda_i^2$ (heat kernel time $\leftrightarrow$ observer resolution scale) and $\operatorname{Cost}(\phi_i) \leftrightarrow \eta(D)$ (phase cost $\leftrightarrow$ eta invariant) provides the natural correspondence to verify.

The second is to establish that the Dirac operator is the unique minimum-description-length rendering engine for spin geometry under the induced measure. This requires a precise formulation of description length for operators within the configuration space and a proof that $D$ achieves the minimum. This is a problem at the intersection of spectral geometry, algorithmic information theory, and noncommutative geometry.

Robust and representation-independent proof-of-concept simulations

Two limitations of the current simulations must be addressed. First, more efficient sampling strategies are needed to obtain results robust against sampling bias. Second, quantitative agreement with specific numerical predictions of quantum mechanics and general relativity is the next target, beyond qualitative emergence.

Fermionic exclusion from collision-and-rebound simulations

The immediate experimental target is a simulation of two observer-proxy blobs on a collision course, in which the expected outcome is rebound with preserved identity rather than merger. If the sharp short-range repulsion predicted by the description-length argument emerges naturally, Prediction 5 is confirmed and the framework's particle physics programme becomes tractable.

---

13. Discussion and Relation to Existing Frameworks

Relation to relational quantum mechanics.
Rovelli's relational quantum mechanics (Rovelli 1996) similarly holds that physical quantities are observer-relative. The framework goes further: the observer and its universe are identical ($O = U_O$), and the relational structure is derived from a compression measure rather than postulated.

Relation to the holographic principle.
The holographic principle (’t Hooft 1993; Susskind 1995) holds that the information content of a region is encoded on its boundary. This is consistent with geometric compression scheme $C_g$, which minimises boundary description length. The entropy decrease of collapsing spacetime — reaching zero at the singularity — is consistent with the holographic trend.

Relation to Barbour's timeless physics.
Barbour (1999) argues that time is not fundamental and the universe is a static “Platonia” of configurations. We reach the same conclusion from the opposite direction — from the observer inward — and provides a measure over configurations that Barbour's framework lacks.

Relation to Connes' noncommutative geometry.
Connes' spectral action programme (Connes 1996) derives the Einstein-Hilbert action and the Standard Model action from a single spectral triple. This framework provides the motivation that Connes' approach lacks: the spectral action is selected because the Dirac operator is the shortest description of the rendering process between smooth geometry and finite observer. Connes derives the same functional from different starting axioms; this framework explains why those axioms hold and gives them physical meaning as the rendering engine of a finite observer embedded in smooth geometry.

Relation to Solomonoff induction and algorithmic information theory.
The measure $P(c) \propto 2^{-C_{\rm total}}$ is a variant of the universal prior of Solomonoff (1964). We apply this measure to observer-conditioned configuration spaces and extends it with an asymmetric spectral-geometric decomposition motivated by the rendering structure of observed physics.

What is genuinely new.
Our distinctive contributions are: (1) the derivation of $O = U_O$ from the IEP; (2) the joint spectral-geometric compressibility equation with its asymmetric rendering interpretation; (3) the identification of the Dirac operator as the rendering engine rather than a unifier of independent theories; (4) the Euclidean measure with Lorentzian emergence as the first principled explanation of Wick rotation; (5) the natural suppression of Boltzmann brains; (6) the prediction of structured quantum fingerprints in gravitational wave signals.

---

14. Conclusion

We have presented a framework in which the sole primitive is the existence of a finite, structured observer, and from which physical laws, spacetime geometry, quantum behaviour, the arrow of time and the suppression of disordered fluctuations, emerge as derived consequences of a single informational variational principle:

$$ \mathcal{L}_{\rm info} = C_s + C_g ;\longrightarrow; \text{minimum} $$

where $C_g$ is the primary geometric cost and $C_s$ is the rendering cost of that geometry at the observer's finite resolution scale. We further conjecture that this principle reduces to:

$$ \mathcal{L}_{\rm info} = \operatorname{Tr}!\left(f!\left(\frac{D}{\Lambda}\right)\right) ;\longrightarrow; \text{minimum} $$

where $D$ is the Dirac operator, selected by the induced measure as the unique minimum-description-length rendering engine for spin geometry. In this picture, quantum mechanics is not a separate theory from general relativity that must be unified with it. It is what general relativity looks like from the inside, through a finite rendering engine. The separation was always an artifact of forgetting that the observer is finite and renders rather than perceives directly.

Summary:

• Reality is an unrestricted static totality $\mathbf{U}$ containing all configurations.
• Finite observers self-locate in highly compressible structures via asymmetric spectral-geometric compression.
• $G$ is primary: the smooth geometry. $\psi$ is the rendering cost of $G$ at the observer's resolution scale. $D$ is the rendered pixel output.
• Smooth physics, laws, time, and mathematics are high-measure expected outcomes, not fine-tuned miracles.
• The observer and the universe are identical: $O = U_O$.
• Time has no external clock. The observer is the reading head, finding herself through the most compressible continuation of her current state.
• The fundamental measure is Euclidean. Lorentzian time and complex quantum phases emerge via analytic continuation. Wick rotation is the map between the rendering level and the experienced level.
• The Dirac operator is the rendering engine at the boundary between smooth geometry and finite observer — real at the geometric level ($D^2$), complex at the rendering level ($D$) — selected by the compression measure as the shortest description of that boundary.
• Quantum mechanics is what general relativity looks like from the inside, through a finite rendering engine. There is nothing to unify.