DeepCubeA, the subject of the study, was published inNature Machine Intelligenceon July 15, 2019.
What makes this AI special is that it learned how to solve the Rubik on its own.
They started with a computer simulation of a completed puzzle and then scrambled the cube.
It learned on its own, Baldi noted.
Our AI takes about 20 moves, most of the time solving it in the minimum number of steps.
This work is a step toward this hefty goal.
source: www.techworm.net
