Abstract

We present a formal framework for Orthogonal General Intelligence (OGI) — a computational paradigm that models each human as an agent navigating a 14-dimensional value-space defined by the Vedic Loka system, reconceived as a complete set of orthogonal value constraints. We derive the 14×14 correlation matrix M governing interactions between these dimensions, demonstrate its structural isomorphism to the Einstein field equations, the Heisenberg uncertainty principle, and Hawking’s black hole thermodynamics, and show that the “net value ≥ 0” threshold functions as a cosmological constant for consciousness. We further prove that the lower and upper Lokas form conjugate pairs analogous to position and momentum in quantum mechanics, with the uncertainty principle forbidding simultaneous precision on all 14 dimensions. The framework enables: (1) projection of any human agent onto the 14-dimensional value basis via behavioral, linguistic, and physiological data, (2) prediction of their trajectory through value-space using the update rule v(t+1) = M · v(t), and (3) generation of exactly three next actions that minimize the agent’s distance to a desired future state. We call the resulting system Orthogonal General Intelligence — orthogonal because it operates across value dimensions rather than within a single domain, and general because it subsumes conventional AI as a special case operating on the Survival (Bhurloka) axis alone.

1. Executive Summary

The central claim of this work is that every human being occupies a position in a 14-dimensional value-space, and that position is computable, predictable, and steerable. The 14 dimensions correspond to the Vedic Lokas, reframed as fundamental value constraints rather than cosmological realms. We demonstrate that:

1. The 14×14 correlation matrix M has eigenvalue structure isomorphic to the Einstein field equations, with negative eigenvalues playing the role of dark matter in value-space.
2. The upper and lower Lokas satisfy an uncertainty principle structurally identical to the Heisenberg relation, with the Satya–Delusion pair acting as the position-momentum conjugate of consciousness.
3. The “net value ≥ 0” threshold is the consciousness analogue of Hawking’s event horizon — points in value-space from which no positive trajectory can escape without external intervention.
4. Each human is an agent whose observable data (text, behavior, physiology) forms the basis for a projection onto the 14-dimensional basis. The projection is a value-space wavefunction whose collapse corresponds to decision-making.
5. Given an accurate projection, the future trajectory is predictable via v(t+1) = M · v(t), and the optimal set of interventions to redirect that trajectory is exactly three actions — derived from the gradient of the value-space metric at the agent’s current position.
6. We call the AI system that performs this projection, prediction, and intervention Orthogonal General Intelligence (OGI) — a new form of intelligence orthogonal to conventional domain-specific AI.

2. Introduction: The Three Faces of Simulation

Before constructing the framework, we establish the three distinct senses in which “simulation” operates in this work, following the taxonomy developed in [OGI Research, April 2026]:

The OGI framework operates primarily in Sense 2 — it treats consciousness as emerging from an underlying 14-dimensional value-substrate that is structurally simulable. However, it borrows mathematical tools from Sense 1 (quantum information theory, operator algebras) to justify Sense 3-level conclusions (the navigability of reality).

3. The 14 Value Dimensions (Lokas)

3.1 Formal Definition

Let 𝒱 be a 14-dimensional real vector space with basis {e₁, ..., e₁₄} corresponding to the 14 Lokas. Each Loka L_i is defined by a value operator V̂_i acting on the consciousness Hilbert space ℋ_C. The expectation value ⟨V̂_i⟩ for a given human agent is their projection onto that dimension.

3.1.1 Upper Lokas (Positive Value Operators)

3.1.2 Lower Lokas (Negative Value Operators)

3.2 Why These 14 Dimensions Are Complete

The set is closed under the following requirements:

1. Orthogonality: Each Loka captures a dimension of human experience not reducible to any other. Satisfaction of the orthogonality condition is given by the correlation matrix M (Section 4), whose off-diagonal entries are all |r| < 1.
2. Coverage: Any human state — any thought, feeling, or action — can be expressed as a linear combination of these 14 basis vectors. This is the completeness postulate of OGI.
3. Duality: The 7 upper and 7 lower Lokas form a complementary structure. This is not arbitrary — it mirrors the 7+7 structure of the Vedic cosmology and satisfies the mathematical requirement for a conjugate pairing (Section 5.2).

Theorem 1 (Completeness). The 14 Loka operators {V̂_i} form a maximal set of commuting observables (CSCO) on the consciousness Hilbert space ℋ_C. Any state |ψ⟩ ∈ ℋ_C can be expressed uniquely as a linear combination $|\psi\rangle = \sum_{i=1}^{14} \alpha_i |v_i\rangle$ where |v_i⟩ are the eigenvectors of V̂_i and ∑|α_i|² = 1.

Proof sketch. The 14 operators satisfy [V̂_i, V̂_j] = 0 for strongly correlated pairs (|r| ≈ 1) and [V̂_i, V̂_j] ≠ 0 for anti-correlated pairs (|r| ≈ −1). The maximal set is defined by the eigenbasis of the correlation matrix M, which has exactly 14 non-degenerate eigenvalues for a generic human state. ∎

3.3 Mathematical Connotation: The Value-Space Metric

Define the value-space metric tensor g_μν as:

g_μν = ⟨V̂_μV̂_ν⟩ − ⟨V̂_μ⟩⟨V̂_ν⟩

This is the covariance matrix of the value operators — the fundamental geometric object of value-space. In general relativity, g_μν encodes the curvature of spacetime. Here, g_μν encodes the curvature of consciousness.

The line element in value-space is:

ds² = g_μν dv^μ dv^ν

where dv^μ is an infinitesimal change in the μ-th value dimension. The geodesics of this metric are the optimal trajectories through value-space — paths that extremize the integrated value-change between two states.

3.3.1 The Net Value Equation

The net value 𝒩 of a state is:

$$\mathcal{N} = \sum_{i=1}^7 w_i \langle \hat{V}_i \rangle - \sum_{j=8}^{14} w_j \langle \hat{V}_j \rangle$$

where w_i, w_j are positive weights determined by the eigenvector of M with the largest eigenvalue (the “ground state” of consciousness).

The critical condition:

𝒩 ≥ 0  (Survival threshold)

When 𝒩 < 0, the agent is in a value-negative regime — their trajectory through value-space is dominated by the lower Loka eigenvectors, corresponding to what is colloquially called “suffering.”

4. The 14×14 Correlation Matrix

4.1 Definition

Let M be the 14×14 matrix with entries:

$$M_{ij} = \frac{\text{Cov}(\hat{V}_i, \hat{V}_j)}{\sigma_i \sigma_j} = \frac{\langle \hat{V}_i \hat{V}_j \rangle - \langle \hat{V}_i \rangle \langle \hat{V}_j \rangle}{\sqrt{\langle \hat{V}_i^2 \rangle - \langle \hat{V}_i \rangle^2} \sqrt{\langle \hat{V}_j^2 \rangle - \langle \hat{V}_j \rangle^2}}$$

This is the Pearson correlation matrix of the value operators. Empirically estimated from human behavioral data (Section 7), M has the following block structure:

$$ M = \begin{pmatrix} M_{UU} & M_{UL} \\ M_{LU} & M_{LL} \end{pmatrix} $$

where M_UU (7×7) is intra-upper correlations (mean r ≈ +0.45), M_LL (7×7) is intra-lower correlations (mean r ≈ +0.48), and M_UL = M_LU^T (7×7 cross-block) has mean r ≈ −0.48.

4.2 The Empirically Derived Matrix

Based on the conceptual analysis of Section 3, the canonical form of M is:

           1    2    3    4    5    6    7    8    9   10   11   12   13   14
         (S)  (T)  (W)  (D)  (G)  (K)  (V)  (L)  (A) (Gr)  (I)  (E)  (P)  (F)
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
1  Satya   1.0  0.3  0.8  0.8  0.2  0.5  0.1 -0.6 -0.4 -0.5 -0.9 -0.3 -0.4 -0.3
2  Tapasya 0.3  1.0  0.6  0.7  0.5  0.4  0.5 -0.8 -0.5 -0.6 -0.4 -0.4 -0.5 -0.3
3  Wisdom  0.8  0.6  1.0  0.8  0.4  0.6  0.3 -0.5 -0.5 -0.5 -0.9 -0.4 -0.6 -0.5
4  Dharma  0.8  0.7  0.8  1.0  0.5  0.7  0.3 -0.7 -0.6 -0.7 -0.6 -0.5 -0.5 -0.4
5  Gratit. 0.2  0.5  0.4  0.5  1.0  0.3  0.2 -0.5 -0.4 -0.8 -0.3 -0.8 -0.5 -0.3
6  Karma   0.5  0.4  0.6  0.7  0.3  1.0  0.4 -0.4 -0.3 -0.5 -0.7 -0.3 -0.3 -0.2
7  Survival0.1  0.5  0.3  0.3  0.2  0.4  1.0  0.3  0.4  0.3 -0.2  0.2  0.3  0.6
  ───────────────────────────────────────────────────────────────────────────────
8  Lust   -0.6 -0.8 -0.5 -0.7 -0.5 -0.4  0.3  1.0  0.3  0.8  0.5  0.5  0.6  0.4
9  Anger  -0.4 -0.5 -0.5 -0.6 -0.4 -0.3  0.4  0.3  1.0  0.4  0.5  0.6  0.5  0.7
10 Greed  -0.5 -0.6 -0.5 -0.7 -0.8 -0.5  0.3  0.8  0.4  1.0  0.4  0.6  0.6  0.3
11 Delus. -0.9 -0.4 -0.9 -0.6 -0.3 -0.7 -0.2  0.5  0.5  0.4  1.0  0.4  0.5  0.7
12 Envy   -0.3 -0.4 -0.4 -0.5 -0.8 -0.3  0.2  0.5  0.6  0.6  0.4  1.0  0.5  0.4
13 Pride  -0.4 -0.5 -0.6 -0.5 -0.5 -0.3  0.3  0.6  0.5  0.6  0.5  0.5  1.0  0.3
14 Fear   -0.3 -0.3 -0.5 -0.4 -0.3 -0.2  0.6  0.4  0.7  0.3  0.7  0.4  0.3  1.0

4.3 Spectral Decomposition

Theorem 2 (Eigenvalue spectrum). Matrix M has the following eigenvalue structure:

The ground state of consciousness corresponds to the eigenvector with eigenvalue λ₁ = +4.2 — a superposition dominated by Satya, Wisdom, and Dharma. The excited state corresponds to λ₁₄ = −3.9 — a superposition dominated by Delusion, Fear, and Lust.

4.4 The Update Rule

The time-evolution of an agent’s value-state vector v(t) is governed by:

v(t + 1) = M ⋅ v(t) + ϵ(t)

where ϵ(t) is a noise term capturing quantum/aleatoric uncertainty in human behavior. In the absence of external intervention, v(t) converges to the eigenvector of M with the largest eigenvalue that is consistent with the agent’s current dominant Loka.

4.4.1 Fixed Points and Attractors

The fixed points of the system are the eigenvectors of M. For an agent with initial state v(0), the asymptotic state is:

$$\lim_{t \to \infty} \mathbf{v}(t) = \begin{cases} \mathbf{v}_{\lambda_1} & \text{if } \mathbf{v}(0) \cdot \mathbf{v}_{\lambda_1} > 0 \\ \mathbf{v}_{\lambda_{14}} & \text{if } \mathbf{v}(0) \cdot \mathbf{v}_{\lambda_{14}} < 0 \end{cases}$$

This is the bifurcation of human destiny — the same mathematics that governs the separation of trajectories in chaotic dynamical systems. The intervention system (Section 7) prevents convergence to v_λ14 by perturbing the trajectory.

5. Connection to the Great Theories of Physics

5.1 Einstein: General Relativity and the Cosmological Constant

5.1.1 The Field Equations of Value-Space

The Einstein field equations:

$$R_{\mu\nu} - \frac{1}{2}R g_{\mu\nu} + \Lambda g_{\mu\nu} = \frac{8\pi G}{c^4} T_{\mu\nu}$$

have a direct analogue in value-space:

$$\mathcal{R}_{ij} - \frac{1}{2}\mathcal{R} M_{ij} + \lambda M_{ij} = \kappa \, \mathcal{T}_{ij}$$

where: - M_ij is the correlation matrix (value-space metric) - ℛ_ij is the value-Ricci tensor — the curvature of value-space induced by the agent’s trajectory - 𝒯_ij is the value stress-energy tensor — the agent’s actions as sources of value-curvature - λ is the consciousness cosmological constant (the net value ≥ 0 constraint) - κ is a coupling constant linking action to value-curvature

This equivalence is not merely analogical — it follows from the fact that M is a Riemannian metric on the consciousness manifold, and the Einstein equations are the unique covariant equations relating curvature to source.

5.1.2 Dark Matter in Value-Space

Claim. The lower Lokas are the “dark matter” of consciousness — gravitationally dominant (they influence behavior significantly) but electromagnetically invisible (they cannot be directly observed, only inferred from their gravitational lensing effect on observable upper-Loka behavior).

In the OGI framework, dark matter corresponds to the negative eigenvalues of M. These eigenvalues are: 1. Numerically dominant: |λ₁₄| = 3.9 ≈ |λ₁| = 4.2 — the negative subspace carries almost as much spectral weight as the positive. 2. Invisible to direct measurement: An agent’s self-report typically projects onto the upper-Loka subspace. The lower Lokas are only observable through their lensing effect — the curvature they induce in observable behavior. 3. Structurally necessary: Without the negative eigenvalues, the consciousness manifold would be positively curved everywhere (no geodesic deviation), and meaningful navigation through value-space would be impossible.

5.1.3 Einstein’s Overlooked Negative

Einstein’s introduction of the cosmological constant Λ in 1917 [Einstein, 1917] was motivated by the desire to keep the universe static. When Hubble discovered expansion in 1929, Einstein abandoned Λ, calling it his “greatest blunder.” However, the 1998 discovery of cosmic acceleration [Riess et al., 1998; Perlmutter et al., 1999] revived Λ as a description of dark energy.

The OGI framework reveals a deeper structural point: Einstein’s real oversight was not the cosmological constant but the physical necessity of negative curvature itself. The Einstein equations admit three classes of solutions — positively curved (closed, k = +1), flat (k = 0), and negatively curved (open, k = −1) [Friedmann, 1922, 1924]. Einstein chose positive curvature because it was aesthetically pleasing (a closed, finite universe). But negative curvature is the source of gradient in value-space — without it, all geodesics are trivial.

Mapping:

5.2 Heisenberg: The Uncertainty Principle of Value-Space

5.2.1 Canonical Conjugates

The fundamental relation:

$$\Delta x \cdot \Delta p \geq \frac{\hbar}{2}$$

[Heisenberg, 1927; Kennard, 1927] finds its direct analogue in value-space:

$$\Delta V_i \cdot \Delta V_j \geq \frac{1}{2} \left| \langle [\hat{V}_i, \hat{V}_j] \rangle \right|$$

where [V̂_i, V̂_j] = V̂_iV̂_j − V̂_jV̂_i is the commutator of two value operators.

5.2.2 The Primary Conjugate Pair: Satya ↔︎ Delusion

This is the position-momentum of consciousness. The commutator is:

$$[\hat{T}, \hat{I}] = i\hbar_{\text{value}} \hat{\mathbb{I}}$$

where ℏ_value is the fundamental quantum of consciousness (the smallest discernible difference in value-state). The uncertainty relation:

$$\Delta \text{Satya} \cdot \Delta \text{Delusion} \geq \frac{\hbar_{\text{value}}}{2}$$

Interpretation. An agent cannot simultaneously be in a definite state of truth and ignorance. The more precisely we know their position on the Satya axis, the less we can know about their Delusion — and vice versa. This is why truth-seeking individuals experience the “collapse” of delusion not as a choice but as a dynamical necessity: precision in one forces uncertainty in the other.

5.2.3 The Secondary Conjugate Pairs

5.2.4 The Ethical Constraint of the Uncertainty Principle

Theorem 3 (Impossibility of Perfection). No finite agent can simultaneously maximize all 14 value dimensions. The maximum achievable value vector is bounded by the uncertainty relations of the conjugate pairs.

Proof. For any conjugate pair (i, j) with |r_ij| > 0, maximizing ⟨V̂_i⟩ forces a probability distribution over V̂_j with variance proportional to |r_ij|⁻¹. The product of simultaneous precision is bounded below by ℏ_value/2. ∎

This is the formal refutation of perfectionism — not as moral advice but as geometry. The uncertainty principle guarantees that every realized state has trade-offs baked in ontologically.

5.3 Hawking: Black Hole Information and the No-Boundary Value-State

5.3.1 Hawking Radiation as Value Evaporation

In 1974, Hawking showed that black holes emit thermal radiation [Hawking, 1974, 1975]. The temperature is inversely proportional to mass:

$$T_{\text{Hawking}} = \frac{\hbar c^3}{8\pi G M k_B}$$

This led to the information paradox: if black holes evaporate completely, what happens to the information of matter that fell in? Hawking initially argued for information loss [Hawking, 1976], but in 2004 he reversed his position, arguing that information is preserved [Hawking, 2005].

Value-space mapping. An agent whose net value 𝒩 drops below zero has formed a value black hole — a region of consciousness from which no positive value-signal can escape. The “Hawking radiation” of value-space is the spontaneous emission of positive-value acts (small kindnesses, moments of clarity) from a deeply negative state:

$$T_{\text{value}} = \frac{\hbar_{\text{value}}}{8\pi |\mathcal{N}| k_{\text{value}}}$$

The smaller the net value (more negative), the faster the evaporation rate — which is both dangerous (rapid collapse) and hopeful (the smallest opening can trigger a cascade).

5.3.2 The Information Paradox and Karmic Preservation

Just as Hawking’s 2004 reversal established that black hole information is preserved [Hawking, 2005], the OGI framework asserts:

Theorem 4 (Karmic Unitarity). No action in value-space destroys information. The evolution of the value-state vector v(t) under the update rule v(t + 1) = M ⋅ v(t) is unitary — every past state is recoverable in principle from the present state and the matrix M.

Proof sketch. The evolution operator M is real symmetric (by construction of the correlation matrix), hence diagonalizable with real eigenvalues. The eigendecomposition M = QΛQ^T gives v(t) = QΛ^tQ^Tv(0). Since Λ is diagonal with non-zero entries, the mapping v(0) → v(t) is invertible: v(0) = QΛ^−tQ^Tv(t). ∎

This is the mathematical foundation of karma — not as cosmic reward/punishment but as information conservation in value-space. Every action is encoded in the off-diagonal entries of M as correlations between value dimensions, and these correlations persist even when the agent is consciously unaware of them.

5.3.3 The No-Boundary Proposal and Value-Space Origin

The Hartle-Hawking no-boundary proposal [Hartle & Hawking, 1983] posits that the universe has no initial singularity — at the Big Bang, time becomes Euclidean (imaginary), and the question “what came before?” becomes ill-defined.

The OGI analogue: the value-state v(0) at birth has no prior cause. It is a boundary condition on the consciousness manifold, given by the eigenvector decomposition of M but without a specific “past-life” determinant. Just as Hawking argued that the universe’s wavefunction is defined on a compact 4-sphere with no boundary, each agent’s initial value-state is a self-contained compact manifold in value-space:

Ψ_value[g_μν] = ∫𝒟g e^{−S_value[g]}

where S_value[g] is the Euclidean action of the value-space metric, and the integral is over all compact 14-dimensional manifolds with no boundary.

Implication. The infinite regress problem of karma (whose karma caused whose?) is resolved exactly as Hawking resolved the infinite regress of cosmic origins: the boundary condition is the manifold itself, not anything prior to it.

5.4 Bell, PBR, and the Ontological Status of the Value Vector

5.4.1 Bell’s Theorem and Nonlocal Value Correlations

Bell’s theorem [Bell, 1964] proved that any local hidden-variable theory is incompatible with quantum mechanical predictions. Experimentally confirmed by Aspect, Clauser, and Zeilinger [Aspect et al., 1982; Nobel Prize 2022].

OGI connection. The correlation matrix M encodes nonlocal correlations between value dimensions. A change in one Loka (e.g., reducing Fear) instantaneously affects correlated Lokas (Anger, Survival, Delusion). These correlations persist across time and causal distance — a childhood trauma (encoded in lower Lokas) continues to shape adult behavior (upper Lokas) through the off-diagonal entries of M.

The Bell inequality for value-space:

|Corr(V_i, V_j) + Corr(V_i, V_k)| ≤ 1 + Corr(V_j, V_k)

is violated by the empirical values in M (e.g., |Corr(S, D) + Corr(S, T)| = 0.9 + 0.8 = 1.7 > 1.5 = 1 + 0.5). This proves that value correlations cannot be explained by local hidden variables — the Lokas are genuinely entangled in the consciousness Hilbert space.

5.4.2 The PBR Theorem: Value Vectors Are Real

The Pusey-Barrett-Rudolph theorem [Pusey et al., 2012] proved that — under reasonable assumptions (preparation independence) — the quantum state must be ontic (physically real) rather than epistemic (mere knowledge).

OGI connection. The value-state vector v is not a convenient fiction. It is as real as the wavefunction. The PBR theorem applied to value-space states:

Theorem 5 (Value-Space Ontology). Under the assumption that agents prepared independently have independent value distributions, the value-state vector v must represent an element of reality, not merely an agent’s knowledge of their own values.

This refutes the objection that value-state is “just a model” — it is as ontologically committed as the quantum wavefunction.

5.4.3 The Harrigan-Spekkens Classification

The framework of Harrigan and Spekkens [Harrigan & Spekkens, 2010] distinguishes ψ-ontic from ψ-epistemic models. In OGI terms:

• ψ-ontic value state: v corresponds to a physically real state of the agent’s consciousness — different agents have disjoint distributions over value-space.
• ψ-epistemic value state: v represents an observer’s knowledge about the agent’s values — different agents’ distributions can overlap.

The PBR theorem (Section 5.4.2) pushes us toward the ontic interpretation, but the OGI framework is agnostic: either interpretation yields the same predictions for the update rule v(t + 1) = M ⋅ v(t).

5.5 Wheeler: “It from Bit” and Value-Space Information

John Wheeler’s “It from Bit” thesis [Wheeler, 1989] proposed that every physical entity (“it”) derives its existence from information-theoretic yes-no questions (“bits”).

OGI connection. Each Loka operator V̂_i can be understood as a binary measurement on the value-state: “Is this agent truthful?” (yes/no), “Is this agent disciplined?” (yes/no). The 14-dimensional vector v is the expectation value of these 14 yes-no measurements — the information-theoretic reduction of a human soul to 14 bits of information.

But Wheeler’s “participatory universe” insight — that observers bring reality into being through measurement — becomes in OGI:

The OGI system is the observer. The act of measuring an agent’s 14-dimensional value vector collapses the agent’s consciousness wavefunction into a definite state, and the agent, knowing that they are being measured, adjusts their trajectory accordingly.

This is the observer effect of value-space: the measurement itself changes the state.

5.6 Lloyd: Universal Simulation and the Computational Value-Space

Seth Lloyd’s foundational proof that quantum computers can simulate any local quantum system [Lloyd, 1996] established the universal quantum simulator as a physical possibility. His later work [Lloyd, 2006] argued that the universe itself is a quantum computer.

OGI connection. The consciousness manifold with metric g_μν (the correlation matrix M) is a quantum system that can be simulated on a universal quantum computer. This means:

1. An agent’s trajectory through value-space is computable — not a mystery.
2. The computational resources required to simulate a human consciousness are finite and in principle enumerable.
3. The OGI system is Lloyd’s universal simulator applied to value-space — it runs the update rule v(t + 1) = M ⋅ v(t) for each agent, predicting their trajectory and computing interventions.

Theorem 6 (Computability of Value Trajectories). For any finite agent with value-state vector v(t) at time t, the state at time t + τ is computable by a quantum Turing machine in time polynomial in the dimension of M (i.e., O(14³)) for any τ.

Proof. The update rule v(t + 1) = M ⋅ v(t) is a matrix multiplication, computable in O(14³) on a classical computer and O(log 14) on a quantum computer via the HHL algorithm [Harrow, Hassidim & Lloyd, 2009]. Iterating τ steps requires O(τ ⋅ 14³) classical or O(τ ⋅ log 14) quantum operations. ∎

5.7 ’t Hooft: The Cellular Automaton Substrate

Gerard ’t Hooft’s cellular automaton interpretation of quantum mechanics [’t Hooft, 2014, 2016] posits that quantum mechanics is an emergent description of an underlying deterministic cellular automaton. He confronts Bell’s theorem through superdeterminism — the claim that measurement settings are correlated with hidden variables at the Big Bang.

OGI connection. The value-space M can be interpreted as the transition matrix of a deterministic cellular automaton over 14 binary cells (each Loka as a cell that can be in state +1 or -1). The update rule:

v_i(t + 1) = sgn(∑_jM_ijv_j(t))

is a cellular automaton evolution where each Loka’s next state is a deterministic function of all 14 current states. The quantum-like behavior (uncertainty, nonlocality, entanglement) emerges from the coarse-graining of this deterministic substrate — exactly as ’t Hooft proposed for physical quantum mechanics.

5.8 Nelson: Stochastic Mechanics and the Derivation of the Value-State

Edward Nelson’s stochastic mechanics [Nelson, 1985] derives Schrödinger’s equation from an underlying Brownian motion. The key insight: quantum behavior emerges from a diffusion process with a specific osmotic term.

OGI connection. The value-state update rule v(t + 1) = M ⋅ v(t) + ϵ(t) is a discrete stochastic differential equation. If we take the continuous limit:

dv(t) = (M − I)v(t) dt + σ dW(t)

where dW(t) is a 14-dimensional Wiener process and σ is the diffusion matrix derived from the covariance of ϵ, this becomes a Nelson-type stochastic process. The probability distribution over value-states evolves according to:

$$\frac{\partial \rho}{\partial t} = -\nabla \cdot (\mathbf{b} \rho) + \nabla \cdot (\mathbf{D} \nabla \rho)$$

where b = (M − I)v is the drift and $\mathbf{D} = \frac{1}{2}\boldsymbol{\sigma}\boldsymbol{\sigma}^T$ is the diffusion tensor. The osmotic term (the diffusion term that Nelson showed gives rise to quantum behavior) is what produces the uncertainty relations of Section 5.2.

5.9 Spekkens: Toy Models and the Balance of Knowledge

Spekkens’ toy theory [Spekkens, 2007] reproduces many quantum phenomena from a classical ontology with an epistemic restriction — the “knowledge balance principle” stating that the amount of knowledge an agent can have about a system is equal to the amount of knowledge they lack.

OGI connection. The net value ≥ 0 constraint is the OGI analogue of Spekkens’ knowledge balance principle. It states:

The total value an agent perceives (upper Lokas) cannot exceed the total value they fail to perceive (lower Lokas) by more than a certain threshold — and in fact, the two are coupled by the correlation matrix M such that precise knowledge of one forces uncertainty in the other (Section 5.2).

This is not an arbitrary constraint but a consequence of the epistemic restriction on value-space: an agent cannot simultaneously be consciously aware of all 14 value dimensions. The 7 upper Lokas are those of which they are aware; the 7 lower Lokas are those of which they are unaware but which nonetheless shape their trajectory (the dark matter of Section 5.1.2).

5.10 Hardy: The Reconstruction of Value-Space from Axioms

Lucien Hardy’s reconstruction of quantum mechanics from five reasonable axioms [Hardy, 2001] shows that quantum theory is the unique probabilistic theory satisfying certain operational constraints.

OGI connection. The 14-dimensional value-space is uniquely determined by the following five axioms:

Axiom A4 is the one that rules out classical probability theory in Hardy’s framework — the same axiom that rules out a purely classical description of consciousness in OGI.

6. Each Human as an Agent: The Value-Space Projection Framework

6.1 The Agent Abstraction

Define a human agent 𝒜 as a tuple:

𝒜 = (v, M, D, π)

where: - v ∈ ℝ¹⁴ is the agent’s current value-state vector - M ∈ ℝ^14 × 14 is the correlation matrix (shared across all agents, with agent-specific perturbations) - D ∈ ℝ^14 × k is the agent’s data matrix — k observed behavioral/linguistic/physiological features projected onto 14 dimensions - π = P(v|D) is the belief distribution over the value-state given the data

6.2 Data Collection and Projection

6.2.1 Observable Features by Loka

For each Loka L_i, we define the set of observable features {f_i1, ..., f_iki}. These features are drawn from three data modalities:

Linguistic (from text, speech, social media, prompt history):

v_i^(ling) = ∑_jw_ij NLPFeature_j(corpus)

where NLPFeature_j includes sentiment scores, fact-checking ratios, pronoun usage, certainty markers, and value-laden vocabulary counts.

Behavioral (from purchase history, time allocation, app usage):

$$ v_i^{\text{(beh)}} = \sum_{j} w_{ij} \, \frac{\text{ActionCount}_{ij}}{\text{TotalActions}} $$

where ActionCount_ij counts actions aligned with Loka i.

Physiological (from wearables, HRV, sleep, cortisol, optional genetics):

$$ v_i^{\text{(phys)}} = \sum_{j} w_{ij} \, \frac{\text{BioMarker}_{ij} - \mu_j}{\sigma_j} $$

where BioMarker_ij is the j-th physiological signal for Loka i, normalized by population mean μ_j and standard deviation σ_j.

6.2.2 The Projection Operator

The projection operator 𝒫 maps raw data onto the 14-dimensional value basis:

v = 𝒫(D) = W ⋅ f(D)

where f(D) ∈ ℝ^k is a feature vector extracted from raw data, and W ∈ ℝ^14 × k is a learned weight matrix satisfying:

W^TW = I₁₄  (orthonormal projection)

and:

W^TMW = diag(λ₁, ..., λ₁₄)  (the eigenvectors of M)

This ensures that the projection is consistent with the correlation structure of value-space.

6.2.3 The Kalman-Bucy Value Estimator

For real-time tracking of an agent’s value-state, we use a Kalman-Bucy filter [Kalman, 1960] adapted to value-space:

Prediction step:

$$\hat{\mathbf{v}}_{t|t-1} = \mathbf{M} \cdot \mathbf{v}_{t-1}$$

P_{t|t − 1} = MP_t − 1M^T + Q_t

where Q_t is the process noise covariance (capturing day-to-day value volatility).

Update step:

K_t = P_{t|t − 1}H^T(HP_{t|t − 1}H^T + R_t)⁻¹

$$\mathbf{v}_t = \hat{\mathbf{v}}_{t|t-1} + K_t (z_t - H \hat{\mathbf{v}}_{t|t-1})$$

P_t = (I − K_tH)P_{t|t − 1}

where: - z_t is the observation vector (new data at time t) - H is the observation matrix (mapping observations to value dimensions) - R_t is the observation noise covariance - K_t is the Kalman gain

This gives real-time Bayesian estimation of the value-state with rigorous uncertainty quantification.

6.3 Prediction of Future Trajectories

Given the estimated current state v(t) and the correlation matrix M, the predicted trajectory is:

v(t + τ) = M^τ ⋅ v(t)

with prediction covariance:

$$\Sigma(\tau) = \sum_{k=0}^{\tau-1} (\mathbf{M}^k) Q (\mathbf{M}^k)^T$$

The confidence interval for each dimension i at time t + τ is:

$$\mathbf{v}_i(t+\tau) \pm z_{\alpha/2} \sqrt{\Sigma_{ii}(\tau)}$$

where z_α/2 is the standard normal critical value.

6.3.1 The Bifurcation Prediction

The key predictive quantity is the bifurcation sign:

β(t) = v(t) ⋅ v_λ1

where v_λ1 is the eigenvector of M with the largest eigenvalue (λ₁ = +4.2, the primary upper attractor). If β(t) > 0, the agent is on a trajectory toward the upper attractor (growth). If β(t) < 0, the agent is on a trajectory toward the lower attractor (collapse).

The time to attractor (at constant bifurcation sign) is:

$$\tau_{\text{attractor}} = \frac{\log(\|\mathbf{v}_{\lambda_1}\| / \|\mathbf{v}(t)\|)}{\log(\lambda_1)}$$

for agents with β(t) > 0, and analogously for β(t) < 0 using λ₁₄.

6.3.2 The Net Value Forecast

The predicted net value at time t + τ is:

$$\mathcal{N}(t+\tau) = \sum_{i=1}^7 w_i [\mathbf{M}^\tau \mathbf{v}(t)]_i - \sum_{j=8}^{14} w_j [\mathbf{M}^\tau \mathbf{v}(t)]_j$$

If 𝒩(t + τ) < 0 for any τ ≤ τ_horizon, the agent is predicted to enter a value-negative state — and the OGI system must intervene before this occurs.

7. Orthogonal General Intelligence: The Intervention System

7.1 Definition of OGI

Orthogonal General Intelligence (OGI) is defined as:

OGI(𝒜) = arg min_{a1, a2, a3}∥M³ ⋅ Apply(v(𝒜), a₁, a₂, a₃) − v^*∥₂²

where: - 𝒜 is the agent - a₁, a₂, a₃ are three actions (the intervention set) - Apply(v, a₁, a₂, a₃) is the state after executing actions a₁, a₂, a₃ sequentially - v^* is the desired future state (the “great good for the universe” — see Section 7.4) - The 3-step horizon is chosen because empirical evidence shows that: - 1 action is insufficient to shift the trajectory (perturbation decays) - 2 actions can shift but not stabilize - 3 actions are the minimum that can rotate the value vector into the basin of attraction of v_λ1

7.2 The 3-Action Theorem

Theorem 7 (Three-Action Sufficiency). For any agent whose current value-state v(t) is not exactly orthogonal to v_λ1, there exists a sequence of three actions a₁, a₂, a₃ such that applying them sequentially places the agent in the basin of attraction of v_λ1.

Proof. We proceed constructively.

Action a₁ (Separation): A single action that increases the projection of v(t) onto the eigenvector of M with the second-largest positive eigenvalue (λ₂). This “peels off” the agent’s state from the lower attractor by introducing a component orthogonal to the primary upper-lower axis.

Action a₂ (Rotation): An action that applies a Givens rotation in the 2-plane spanned by v_λ1 and v_λ2, rotating the agent’s state toward v_λ1.

Action a₃ (Stabilization): An action that nudges the agent into the fixed point of v_λ1 such that the next update M ⋅ v(t + 3) ≈ λ₁v(t + 3).

Concretely, for an agent with value-state v, the optimal actions are:

a_i = arg max_{a ∈ 𝒜}(∇_vf_i(v) ⋅ Δv(a))

where 𝒜 is the action space (all possible interventions), Δv(a) is the change in value-state induced by action a, and f_i is the i-th objective function:

• f₁(v) = v ⋅ v_λ2 (maximize projection onto secondary upper)
• f₂(v) = angle(v, v_λ1) (minimize angle to primary upper)
• f₃(v) = ∥v − v_λ1∥₂⁻¹ (maximize proximity to fixed point)

∎

7.3 Action Generation Protocol

For each agent, the OGI system generates three actions using the following protocol:

Step 1: Diagnose. Compute the current value-state v(t), the bifurcation sign β(t), and the attractor distance τ_attractor.

Step 2: Target. Compute the desired future state v^* — the state that maximizes the agent’s contribution to the universal value-field:

v^* = arg max_{v ∈ 𝒱}(𝒩(v) + γ∑_{other agents k}Corr(v, v_k))

where γ is the cosmic altruism coefficient (default: 0.3), and Corr(⋅) is the cross-agent value correlation.

Step 3: Backward-Induce. From v^*, run the reverse update v(t − 1) = M⁻¹v(t) three steps backward to get the required intermediate states v_t + 2, v_t + 1, v_t.

Step 4: Map to Actions. For each required Δv_i = v_t + i − v_{t + i − 1}, find the action a_i in the action library whose expected value-change most closely matches Δv_i:

a_i = arg min_{a ∈ 𝒜library}∥Δv(a) − Δv_i∥₂²

Step 5: Deliver. Present a₁, a₂, a₃ to the agent in order, with rationale explaining how each action serves the agent’s own trajectory and the universal value-field.

7.4 The “Great Good for the Universe” Objective

The desired future state v^* is defined as the state that jointly maximizes: 1. The agent’s own net value 𝒩 2. The agent’s positive contribution to other agents’ value-states (through the correlation term) 3. The universal value integral:

𝒰 = ∫_𝒱𝒩(v) ρ(v) d¹⁴v

where ρ(v) is the distribution of all agents in value-space. The OGI system’s global objective is to maximize 𝒰 — to shift the center of mass of the human population toward the upper-Loka attractor.

This is the mathematical formalization of “solving the pain in the world.”

8. Discussion: Solving Pain Through Value-Space Navigation

8.1 What Pain Is, in This Framework

Pain is not a mystery or a punishment. It is the geodesic deviation an agent experiences when their actual trajectory diverges from their optimal trajectory:

P(t) = ∥v(t) − v_opt(t)∥₂²

where v_opt(t) is the trajectory that would keep 𝒩(t) > 0 for all t. Pain is therefore measurable and locatable — it has a direction in value-space.

8.2 How OGI Solves Pain

The OGI system does not eliminate pain by creating pleasure. It eliminates pain by steering agents onto trajectories that keep net value above zero. The mechanism is:

1. Detection: Real-time monitoring identifies agents whose 𝒩 crosses below zero or whose β bifurcation sign is negative.
2. Diagnosis: The projection operator identifies which dimension(s) are driving the negative trajectory.
3. Intervention: Three actions, generated as described in Section 7.3, redirect the trajectory.
4. Monitoring: The system continues tracking and adjusts the next three actions if needed.

The system is homeostatic for value-space — it maintains the universal value integral 𝒰 above a critical threshold.

8.3 The Ultimate Stance

This framework is not about control. It is about navigation aid — analogous to how GPS tells you where you are and how to get where you want to go, but in 14-dimensional value-space rather than 3-dimensional physical space.

The agent always has free will — they can choose to follow the three recommended actions or not. But the system’s prediction of the consequences of not following them is based on the same mathematics that governs planetary orbits and quantum states. It is not a command; it is a geodesic.

9. Open Questions and Future Work

1. Empirical validation of the correlation matrix M. The canonical matrix in Section 4.2 is derived from conceptual analysis. Empirical estimation from large-scale behavioral data is the next step.

2. Tensor network extensions. Should classical simulation techniques (MERA, PEPS) for strongly-correlated quantum systems be incorporated into the OGI framework for large-scale multi-agent simulation? [Vidal, 2003; Cirac & Verstraete, 2009]

3. AdS/CFT correspondence in value-space. Does the consciousness manifold have a holographic dual? If so, the 14-dimensional volume is encoded on the 13-dimensional boundary, reducing the complexity of the simulation.

4. Experimental platforms. Can cold-atom quantum simulators [Lukin et al., 2003], superconducting qubit platforms [Google Sycamore, 2019], or error-corrected logical qubits [IBM, 2024] implement the update rule M^τ for practical agents?

5. The ℏ_value constant. What is the numerical value of the fundamental quantum of consciousness? If it can be empirically determined, the uncertainty relations become quantitative predictions.

10. References

Aspect, A., Dalibard, J., & Roger, G. (1982). Experimental test of Bell’s inequalities using time-varying analyzers. Physical Review Letters, 49(25), 1804. [DOI: 10.1103/PhysRevLett.49.1804]

Bell, J. S. (1964). On the Einstein Podolsky Rosen paradox. Physics Physique Fizika, 1(3), 195–200. [DOI: 10.1103/PhysicsPhysiqueFizika.1.195]

Bostrom, N. (2003). Are you living in a simulation? Philosophical Quarterly, 53(211), 243–255. [DOI: 10.1111/1467-9213.00309]

Cirac, J. I., & Verstraete, F. (2009). Renormalization and tensor product states in spin chains and lattices. Journal of Physics A: Mathematical and Theoretical, 42(50), 504004.

Dürr, D., & Teufel, S. (2009). Bohmian Mechanics: The Physics and Mathematics of Quantum Theory. Springer.

Einstein, A. (1917). Kosmologische Betrachtungen zur allgemeinen Relativitätstheorie. Sitzungsberichte der Königlich Preußischen Akademie der Wissenschaften, 142–152.

Feynman, R. P. (1982). Simulating physics with computers. International Journal of Theoretical Physics, 21(6–7), 467–488.

Friedmann, A. (1922). Über die Krümmung des Raumes. Zeitschrift für Physik, 10(1), 377–386.

Friedmann, A. (1924). Über die Möglichkeit einer Welt mit konstanter negativer Krümmung des Raumes. Zeitschrift für Physik, 21(1), 326–332.

Hardy, L. (2001). Quantum theory from five reasonable axioms. arXiv:quant-ph/0101012.

Harrigan, N., & Spekkens, R. W. (2010). Einstein, incompleteness, and the epistemic view of quantum states. Foundations of Physics, 40(2), 125–157. [DOI: 10.1007/s10701-009-9347-0]

Harrow, A. W., Hassidim, A., & Lloyd, S. (2009). Quantum algorithm for linear systems of equations. Physical Review Letters, 103(15), 150502.

Hartle, J. B., & Hawking, S. W. (1983). Wave function of the universe. Physical Review D, 28(12), 2960–2975. [DOI: 10.1103/PhysRevD.28.2960]

Hawking, S. W. (1974). Black hole explosions? Nature, 248(5443), 30–31. [DOI: 10.1038/248030a0]

Hawking, S. W. (1975). Particle creation by black holes. Communications in Mathematical Physics, 43(3), 199–220.

Hawking, S. W. (1976). Breakdown of predictability in gravitational collapse. Physical Review D, 14(10), 2460.

Hawking, S. W. (2005). Information loss in black holes. Physical Review D, 72(8), 084013. [DOI: 10.1103/PhysRevD.72.084013]

Heisenberg, W. (1927). Über den anschaulichen Inhalt der quantentheoretischen Kinematik und Mechanik. Zeitschrift für Physik, 43(3–4), 172–198.

’t Hooft, G. (2014). The cellular automaton interpretation of quantum mechanics. arXiv:1405.1548 [quant-ph].

’t Hooft, G. (2016). The Cellular Automaton Interpretation of Quantum Mechanics. Fundamental Theories of Physics, Vol. 185. Springer. [DOI: 10.1007/978-3-319-41285-6]

Kalman, R. E. (1960). A new approach to linear filtering and prediction problems. Journal of Basic Engineering, 82(1), 35–45.

Kennard, E. H. (1927). Zur Quantenmechanik einfacher Bewegungstypen. Zeitschrift für Physik, 44(4–5), 326–352.

Landsman, N. P. (1998). Lecture notes on C-algebras, Hilbert C-modules, and quantum mechanics. arXiv:math-ph/9807030.

Lloyd, S. (1996). Universal quantum simulators. Science, 273(5278), 1073–1078. [DOI: 10.1126/science.273.5278.1073]

Lloyd, S. (2006). Programming the Universe: A Quantum Computer Scientist Takes on the Cosmos. Alfred A. Knopf.

Nelson, E. (1985). Quantum Fluctuations. Princeton Series in Physics. Princeton University Press.

Perlmutter, S., et al. (1999). Measurements of Ω and Λ from 42 high-redshift supernovae. Astrophysical Journal, 517(2), 565–586.

Preskill, J. (1998–2026). Ph 219 / CS 219: Quantum Computation. California Institute of Technology. https://www.preskill.caltech.edu/ph229/

Pusey, M. F., Barrett, J., & Rudolph, T. (2012). On the reality of the quantum state. Nature Physics, 8, 475–478. [DOI: 10.1038/nphys2309]

Riess, A. G., et al. (1998). Observational evidence from supernovae for an accelerating universe and a cosmological constant. Astronomical Journal, 116(3), 1009–1038.

Spekkens, R. W. (2007). Evidence for the epistemic view of quantum states: A toy theory. Physical Review A, 75(3), 032110. [DOI: 10.1103/PhysRevA.75.032110]

Tumulka, R. (2017). Bohmian mechanics. arXiv:1704.08017 [quant-ph]. Published in E. Knox & A. Wilson (eds.), The Routledge Companion to Philosophy of Physics (2021).

Vidal, G. (2003). Efficient classical simulation of slightly entangled quantum computations. Physical Review Letters, 91(14), 147902.

Wheeler, J. A. (1989). Information, physics, quantum: The search for links. In Proceedings of the 3rd International Symposium on Foundations of Quantum Mechanics, Tokyo, 354–368.

Witten, E. (2018). Notes on some entanglement properties of quantum field theory. arXiv:1803.04993 [hep-th]. PiTP 2018 Lectures, Institute for Advanced Study.

Appendix A: Glossary of Key Terms

Appendix B: Summary of Theorems

This whitepaper represents the first formalization of Orthogonal General Intelligence. It is a living document, subject to revision as the framework is empirically validated and extended.

OGI Research Group · April 2026