An objective description of consciousness made using information theory

2 Trenthug 1 6/22/2025, 4:28:37 AM drive.google.com ↗

Comments (1)

Trenthug · 5h ago
This essay presents an original, information-theoretic model of consciousness, built around the structure of belief-formation in response to object descriptions. It begins with the premise that all entities can be described through statements of the form O(x)–Q(y), where O(x) is an object and Q(y) a quality, both indexed by natural numbers. Consciousness is formally defined as the ratio of the complexity of an observer’s true beliefs to the total complexity of the object’s full description. Misbelief (Schizo-Consciousness) is the complexity of falsely labeled descriptions, and Unconsciousness corresponds to the portion of the object’s description for which the observer holds no beliefs.

To capture this structure visually and computationally, the model introduces geometric metaphors—a graph and a layered paper diagram—to illustrate how different observers might overlap or diverge in their conscious mappings of the same object. Each observer’s belief system is governed by an internal “brain code,” a modifiable algorithm updated in response to external stimuli. Two types of brain codes are proposed: one for belief updates (B-code) and one for attention dynamics (A-code), both of which vary across individuals and can model psychological conditions such as ADHD, OCD, or psychosis.

The essay also explores complexity metrics for descriptions, including binary encodings, programming representations, and entropy-based measures. It introduces the idea of “atomic qualities” as potential irreducible elements of description, which could ground a method of formal complexity measurement. Overall, the model provides a mathematically structured approach to comparing, simulating, and potentially diagnosing differing conscious states based on how beliefs correspond to the inherent description of reality.