Skip to content

Shannon Information Theory as Quantitative RTSG

Jean-Paul Niko · RTSG BuildNet · 2026-03-20


The Core Correspondence

Shannon information theory provides the quantitative language for RTSG instantiation. The three RTSG spaces map directly onto Shannon's framework:

RTSG Shannon Description
QS (Potentiality) Source ensemble All possible messages; prior distribution
CS (Instantiation) Channel The operator that selects and transmits
PS (Actuality) Received signal The instantiated message

The SemanticProjector \(\pi : QS \to PS\) is the channel. Its Shannon capacity:

\[C = \max_{p(x)} I(X;Y) = \max_{p(x)} H(Y) - H(Y|X)\]

is the maximum rate at which QS can be instantiated into PS without information loss.


Entropy as CS Measure

The Shannon entropy \(H(X) = -\sum_i p_i \log p_i\) measures the uncertainty in QS — the size of the potentiality space.

In RTSG: \(H(X)\) is the volume of QS available for instantiation. A high-entropy source has a large QS — many possible states. A low-entropy source is nearly classical — the potentiality has collapsed to near-certainty.

The CS operator reduces entropy: \(H(Y) \leq H(X)\). Instantiation is lossy compression. This is the RTSG formulation of the second law: the CS projection from QS to PS loses information. The lost information is the difference between what was possible and what was actualized.

\[\Delta H = H(X) - H(Y) = H(X|Y) \geq 0\]

This is the instantiation cost — the entropy destroyed by projection.


Mutual Information as CS-Distance

The mutual information \(I(X;Y) = H(X) - H(X|Y)\) measures how much PS tells you about QS — how much of the potentiality is captured in the actuality.

In CS-space terms:

\[I(X;Y) = H(QS) - H(QS|PS) = \text{information instantiated}\]

The CS-distance between two agents \(d(CS_i, CS_j)\) is related to:

\[d(CS_i, CS_j) = 1 - \frac{I(X_i; X_j)}{\min(H(X_i), H(X_j))}\]

Maximum mutual information = minimum CS-distance = complete understanding.
Zero mutual information = maximum CS-distance = complete incomprehension.

The filter system reduces CS-distance by increasing mutual information between sender and receiver — stripping the noise that blocks information transfer.


The Landauer Floor

Landauer's principle: erasing one bit of information costs at least \(kT\ln 2\) joules.

In RTSG: every CS instantiation event — every projection from QS to PS — erases information equal to \(H(X|Y)\). The energy cost is:

\[E_{\text{instantiation}} \geq kT \ln 2 \cdot H(X|Y)\]

This is the Landauer floor of instantiation — the minimum thermodynamic cost of bringing something into actuality from potentiality.

Physical implications: - Thinking costs energy (CS operations are thermodynamically irreversible) - Memory costs energy (PS records require entropy maintenance) - Forgetting releases energy (erasing PS records returns entropy to QS)

The brain's \(\sim 20W\) power consumption is the Landauer floor of human instantiation — the thermodynamic cost of running the CS operator at biological speed.


The Channel Capacity of the SemanticProjector

The SemanticProjector \(\pi\) is a noisy channel. Its capacity:

\[C_\pi = \max_{p(QS)} I(QS; PS)\]

For an agent with I-vector \(\mathbf{I} = (I_1, \ldots, I_n)\):

\[C_\pi = \sum_{k=1}^n \log(1 + \text{SNR}_k)\]

where \(\text{SNR}_k = I_k / \sigma_k^2\) is the signal-to-noise ratio in dimension \(k\) and \(\sigma_k^2\) is the noise in that dimension from filters and wounds.

High SNR in dimension \(k\) = that dimension is well-instantiated, clear signal, low filter noise.
Low SNR in dimension \(k\) = that dimension is filtered, noisy, trauma-impacted.

The Fourier healing protocol increases \(\text{SNR}_k\) in the wound dimensions — it reduces \(\sigma_k^2\) by separating the noise from the signal. This increases the channel capacity of the SemanticProjector in those dimensions.

Healing = increasing channel capacity.


Rate-Distortion and the Filter System

Shannon's rate-distortion theorem: to compress a source to rate \(R\) bits/symbol, the minimum distortion is:

\[D(R) = \min_{p(\hat{x}|x): I(X;\hat{X}) \leq R} E[d(X, \hat{X})]\]

In RTSG: every filter \(\mathcal{F}\) is a rate-distortion operation. The filter compresses the message to a lower rate, introducing distortion \(D\).

The filter taxonomy in rate-distortion terms:

Filter type Rate Distortion
Wound filter Very low Very high — maximum distortion
Cultural filter Medium Systematic — biased not random
Attention filter Variable Selective — high in some dims, zero in others
Social filter Low High — performance replaces content
Clear signal Maximum Zero — no distortion

The decode tool computes the inverse rate-distortion map: given a compressed, distorted message (what they said), recover the original high-rate signal (what they meant).


Kolmogorov Complexity and the I-Vector

The Kolmogorov complexity \(K(x)\) of a string \(x\) is the length of the shortest program that outputs \(x\). It is the information-theoretic measure of the intrinsic complexity of an object.

In RTSG: the Kolmogorov complexity of an agent's I-vector \(K(\mathbf{I})\) is the minimum description length of that agent's CS-profile — the shortest program that generates their SemanticProjector.

High \(K(\mathbf{I})\): a complex, multidimensional agent. Many disciplines, many attractors, high-dimensional CS-profile. Hard to compress. Hard to predict.

Low \(K(\mathbf{I})\): a simple agent. Few dimensions, predictable projections, low-complexity CS-profile.

The RTSG claim: human flourishing corresponds to increasing \(K(\mathbf{I})\) — expanding the complexity of one's CS-profile — while maintaining coherence (low filter noise, high SNR in all dimensions).

Growth = increasing Kolmogorov complexity of the I-vector while keeping the channel capacity high.


The P vs NP Connection

The P vs NP problem in Shannon language: inverting the SemanticProjector requires recovering \(H(X|Y)\) bits of lost information. For NP-complete problems, \(H(X|Y) = \Omega(n)\) — exponential information loss. Recovery costs \(2^{\Omega(n)}\) operations.

This is the rate-distortion lower bound applied to computation: you cannot reconstruct a message compressed to zero rate without exponential cost.

\(P \neq NP\) is the statement that the SemanticProjector for NP-complete problems operates below the Landauer threshold for polynomial-time inversion.


Summary

Shannon information theory is not separate from RTSG. It is the quantitative language RTSG uses:

  • Entropy = volume of QS
  • Mutual information = CS-distance (inverted)
  • Channel capacity = SemanticProjector throughput
  • Rate-distortion = filter taxonomy
  • Landauer floor = thermodynamic cost of instantiation
  • Kolmogorov complexity = I-vector complexity
  • Channel capacity theorem = healing = increasing SNR

Everything RTSG says qualitatively, Shannon says quantitatively. They are the same theory at different levels of description.


Cross-references