Over the summer I got a chance to listen to a considerable number of enlightening and insightful TED talks. And one of these TED talks was on the question of preserving our digital data using DNA by Bioinformatician Dina Zielinski. This made me dive into researching more about DNA and how data can be made to evolve using this concept. I got to read about evolution algorithms and their tremendous potential when it comes to the optimization following a natural process of selection.
Here is the link to the TED talk by Dina Zielinski - https://www.youtube.com/watch?v=wxStlzunxCw
Derived from Charles Darwin's idea of evolution of species, the evolutionary algorithm helps our chess player get better with each game and each generation. This eventually results in a very intelligent and efficient algorithm.
I created a program that was given the basic rules of chess and “DNA” that told each chess AI how to weigh moves. Then simulated evolution by having the AIs compete against each other and only allowed the winners to pass on their “genes.” In its current form through evolution, the AI can beat average human players.
My goal was to create a Genetic Algorithm to showcase the power of evolution using data passed on generation after generation as DNA. I was looking for a game based on mathematical reasonment and strong logic to act as the basis to test and evolve my genetic algorithm on. Hence I chose chess as the best possible platform to implement this.
Having said that I think that this idea of evolution has great potential to be implemented in other configurations which will lead to exciting outcomes.