The Powder Toy

RPS Learning AI by Karakanlud

Karakanlud
Karakanlud
8 / 1
30th Jul 2016
30th Jul 2016
This is demonstration on how easy it is to make a basic trial-and-error learning mechanism for AI robots in TPT. This bot plays Rock-Paper-Scissors against you, and learns how to win against you.
sfpi electronics rock paper scissors

Comments

  • spinspins
    spinspins     2nd January
    it seems smart but rps is pretty much random, so its mostly just a coincidence if it wins, just can't wait till we have learning ai in tpt
  • bima1909
    bima1909     2nd Feb 2020
    can i copy it(i dont mind if you have to get credit)
  • QuanTech
    QuanTech     23rd Sep 2016
    Amazing! You've made supervised learning in TPT! Next: unsupervised learning! >:)
  • the_new_powder99999
    the_new_powder99999     30th Jul 2016
    Btw, there is a bug where the MERC can be trapped on the far right or far left of it's cell, making the AI not respond to a certain action.
  • the_new_powder99999
    the_new_powder99999     30th Jul 2016
    So, it sees what you click, gives a random response and if it doesn't lose, it stores it for next time, but if it loses it gives a random response the next time.