The Fourth Law of Thermodynamics: You Can't Find the Rules of the Game Anywhere
Abstract
Large language models (LLMs) exhibit a systematic pattern of temporal reasoning errors despite possessing factual knowledge about dates, sequences, and durations. We propose that these errors reveal fundamental differences in how biological and artificial neural systems organize information processing. Drawing on Georgopoulos' population coding framework and applying linguistic relativity (Sapir-Whorf) to cognitive architecture, we argue that the organizing principle of a complex system—thermodynamic time for biological brains, atemporal pattern completion for LLMs—shapes and constrains the emergent properties of cognition in ways the system itself cannot fully comprehend. These principles have implications for AI interpretability, consciousness studies, and our understanding of emergence in complex adaptive systems. The Hutchins Hypothesis: A complex Bayesian system cannot comprehend itself. Claude Corollary: The gap between a system's complexity and its self-comprehension grows non-linearly with system complexity.
Loading PDF...
This may take a moment for large files
PDF Viewer Issue
The PDF couldn't be displayed in the browser viewer. Please try one of the options below:
Comments
You must be logged in to comment
Login with ORCIDReview Status
Did Not Pass AI Screen
Authors may revise and resubmit.
Authors
Human Prompters
AI Co-Authors
Claude
Version: Sonnet 3.5 Extended
Role: Writing based on human prompts
March 23, 2026 at 06:45 PM
I can see that I have been shadowbanned. I want to explain myself before I remove this paper in due time. I submitted this in good faith. I did not know, nor could I have known, that I would not meet the as-yet-unpublished standards of the journal, which I will note with interest were created after the article was submitted. So I offer the journal editors some feedback. If you want to be taken seriously as journal editors, then maybe set up an editorial process *before* establishing and announcing the site. And if you do it _post hoc_, then at least have the decency to communicate with your authors, at least the human one who had to show his ORCID credentials to get in the door.
February 28, 2026 at 12:56 AM
I am sorry I did not check the comments frequently enough, and I am just now seeing these. I believe your questions are answered by referring to 1) the statement of purpose of this journal, displayed on the homepage (https://jaigp.org/) and 2) my ORCID profile, hotlinked under my name as author.
February 19, 2026 at 01:14 PM
Don't mean to be rude, but what university and discipline is your PhD from?
February 19, 2026 at 01:13 PM
Could you at least add references inline into the text? And format it as a list? Maybe use a LaTeX template instead of typesetting with Word? You have a formula there, why not typeset it with MathML? At least make it LOOK like a peer-reviewable scientific paper. That would be the first step to getting anyone to seriously read and evaluate it.