>>725181762
>Now memorizing every possible board state isn't possible for a human but for a program, not even AI, it's trivial.
Ok but I have to specify that it's not that deep either, it's even simpler!
Picrel is the structure is used to predict moves in a chess games. Now that's jsut a binary tree but usually for chess it has more branches on each circle.
Basically the program tries to guess the outcome from its next move and for each of your remaining pieces in the field it calculates the best possible outcome following that move, then from that it spawns other moves and redos the predictions based on that previous board states, then it picks the tree branch that has the best possible outcome for him.
Diffculties in such programs is jsut a configuration slider to tell the program how many tree branches it can span at each move (more branches = can guess more possible outcomes = uses more RAM)

>>725182010
I'm a programmerfag. you can learn trees in a month. Trees in programming has absolutely 0 to do with modern AIs and LLMs.