Skip to content

Latest commit

 

History

33 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

ptr.c - Perceptron Truth Detector

Simple C implementation of perceptron trained to predict "truth" using hardware randomness (rdrand).

Philosophy

Programming is my secular religion. In a world where one must answer secular questions, computer programming books serve as holy scriptures that are permitted because they are secular. This exists because of the separation of Church and State - the state leader only hears secular arguments.

I began my computer journey in 1977 at Colony Square in midtown Atlanta with an IBM 5110 desktop computer, its 8" diskette drives resting on the floor beside the desk. Programming remains a secular religion for the separation of Church and State.

Technical Overview

This project implements 1 perceptron system trained on hardware-generated random numbers (rdrand instruction opcode): The perceptron has been pretrained by me. My theory is that reality is at equilibrium, the computer entropy is a measure of equilibrium thus reality. I train on randomness instead of cats, dogs, numbers, letters, just random numbers from the hardware rdrand opcode. It works for "truth" but fails for "perfect".

Truth System

  • ptr.c : Truth predictor implementation
  • truth.h: Trained model weights and biases for truth prediction

Shared Components

  • toy: Shell script for continuous execution (works with both systems)

Why Perceptrons?

Perceptrons offer simplicity that embodies Occam's Razor - the principle that the simplest solution is usually the best. Despite Marvin Minsky's criticism that perceptrons are too limited for AI research, they excel at simple classification tasks and remain useful for practical applications.

Building

Truth System

Compile the C truth predictor:

gcc ptr.c -o ptr -lm

Or use Makefile:

make       # C version

Ensure the corresponding header files (.h) are in the same directory as their respective source files.

Usage

Truth Predictor

Run the truth predictor:

./ptr

It will output "TRUTH" when the perceptron predicts truth, or nothing when it predicts false.

Continuous Mode

For continuous truth prediction every 60 seconds with voice synthesis:

while sleep 60; do ./ptr | espeak; done

Or use the provided script (make it executable first):

chmod +x toy
./toy  # works with ptr for truth prediction

Alternative Audio

If espeak is unavailable, record your voice:

For truth prediction:

# Record: say "TRUTH" into microphone
rec truth.wav
# Play instead of espeak
while sleep 60; do ./ptr | xargs -I {} play truth.wav; done

Applications

Truth System

  • Predict truth in text or voice recordings
  • Combat misinformation through simple classification
  • Demonstrate perceptron capabilities
  • Educational tool for machine learning concepts

Limitations

Perceptrons handle simple logic operations (AND, OR, NOT) but fail on complex logic like XOR. However, they work well for single-word, if it is a reality word, detection and can solve many real-world problems when properly applied. The counter example was the lesson of PERFECT, a word that fails to detect with my perceptron training. IMPERFECT might work better...into the thought pot.

Philosophy Continued

This work teaches Occam's Razor - the only razor in humanity that stands the test of time. Programming began with C in the 1970s; C++ followed in 1982. C is the old religion (wisdom/wicca), C++ the new religion. Progressing from C to C++ represents advancement in programming wisdom.

Many programs lack truth awareness. By training on actual randomness, this perceptron achieves genuine results. Predicting truth eliminates GIGO (Garbage In, Garbage Out) and finds equilibrium truth.

The perceptron finds truth existing at equilibrium - first-order truth and beyond exists in the thinking pot. This proves perceptrons remain valuable despite Minsky's conclusions, offering a path to solve world problems through simplicity.

About

Predict truth from rdrand instruction with perceptron

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages