Deep Thought

An artificial intelligence based on 5-dimensional quantum selection.


The payload

  1. Make an random bit array (e.g. 1101...)
  2. Convert the bits to a string
  3. Process the resulting string
  4. Output it to console


On the start, the script waits for user text input. The input is a question that the Deep Thought is being asked about.


No automatic training

The neural network does not use machine learning algorithms. Thus, the learning process must be done manually by each user, as each person’s perception is different.

Training file format

A valid training file should be formatted as a JSON structure. However, training files parser does not rely on extensions and etc. All values in the file should be in the following format:
"{found_text}": "{output_message}", where {found_text} — substring that answer parser should find in an answer of neuronet, {output_message} — message that answer parser outputs as a message from the Deep Thought.

The default training file is ./{script_name}.trn, where: {script_name} — the filename of script without extension. Usually it is, hence it will look for deep-thought.trn in the same directory as the script.

If the given file path is /dev/null, then it will just clear the training list.


Copyleft of Freedom Level 4 🄯 . Licensed under UNLICENSE license.


View Github