Skip to content

Latest commit

 

History

History
18 lines (9 loc) · 1.18 KB

README.md

File metadata and controls

18 lines (9 loc) · 1.18 KB
Raw

#include<AI>

This is a new, custom, hand-written implementation of a Zero-Knowledge Reinforcement Learning algorithm that, unlike AlphaZero, does NOT suffer from the shallow trap problem and should be able to play solved games perfectly.

Marvel over this stb-style #include-only single-header library focusing on General Game Playing, aka. Effective Decision Making under Uncertain or Incomplete Information (EDMUII). It aims to implement Artificial General Intelligence as a single-header drop-in library with an emphasis on quick deployment and ease-of-use. General Game Playing is the most difficult1 problem in the universe, far surpassing quantum physics, elliptic curve arithmetic and axionic singularity modulation. As such, it remains an ongoing research project! Thread carefully 🔬

Building

  1. Include includeai.hpp
  2. Inspect the examples (in the example folder) on how to use
  3. Profit...

Footnotes

  1. While still falling towards the "somewhat possible" side on the possible/impossible dividing line... ↩