Ai in Scratch! (Artificial Intelligence)
Game Introduction
Is this ethical? How does this work? How can a program learn? These are all great questions. This is an evolutionary ai, which means that the mistakes that lead to losing are pruned off each time. (Negative reinforcement) However, when the ai wins, it reinforces that behavior so it's more likely to choose to win again. (Positive reinforcement) But, is this ethical? Well, the human brain has 86 billion neurons (86,000,000,000). A worm has 302 neurons. This ai has about 84 neurons on a good day. (It makes a neuron for each possible action when it learns it, so the number changes) Also, this ai is highly specialized. Its fed possible moves, and it chooses one based off of its past experience. You can think of it as a fancy program.
How To Play
You are in charge of 3 blue pawns. They can move forward a step, or diagonal to capture an opposing pawn. (Click on a pawn to see its moves. Click on the previewed move to choose it!) You win by either getting one of your pawns to the opposite side of the board, or blocking the opponent from moving their pawns. Each round you play, the opposing player learns from yours and its mistakes. However, if you don't wish to take the time to train the ai, you can hit the fast training button in the bottom left corner to give it rapid random matches. #ai the ai’s code is created by me, the original 3 by 3 chess game is called hexapawn. More info about hexapawn can be found here! https://en.m.wikipedia.org/wiki/Hexapawn
Author
minekraftkid
Category
Game Information
Game Popularity
4.8k views
Collection Count
327 favorites
