Why write a program that will play Chess itself?

Question

Python

Passionate about chess. I got the idea to write my own bot that would be a worthy opponent for me, and ideally would even beat me.
Where to start writing this bot??
I know Python plus or minus well. PyQT also used in his crafts.
But I've never dealt with machine learning.
If no water, what do I need?
1) Write Chessboard object classes?
2) Mark the "field of coordinates" itself, that is, the board itself.
3) Give each piece its move algorithm.
well, and then everything... it's not a damn thing how to teach the program tactics, at least the basic one, how to set its goal - to checkmate the opponent's king?

Answer 1

From reading the theory and drawing up a mathematical model and a mathematical algorithm.
In general, the topic of chess bots is hackneyed and well disclosed. Machine learning is generally not needed here.

Answer 2

There is such a neural network "Reinforcement learning" (I would call it "carrot and stick") - Reinforcement Learning.
You can do as the AI ​​learned from the game of Go: one program plays with another and everyone helps each other tune the neural network.
An example from Matlab, how a robot was taught to walk .