Scientist Stephen McAleer and a group of artificial intelligence researchers have developed an AI called ‘DeepCube’ that can solve a Rubik’s Cube without human assistance.
In a report by Techspot.com, McAleer and the team discussed in their paper titled “Solving the Rubik’s Cube Without Human Knowledge” how a technique called “Autodidactic Iteration” allows a deep-learning computer to teach itself how to solve a Rubik’s Cube.
McAleer said that DeepCube can solve the Rubik’s Cube 100 percent of the time from any scrambled state without human interaction.
He said: “Our algorithm is able to solve 100% of randomly scrambled cubes while achieving a median solve length of 30 moves—less than or equal to solvers that employ human domain knowledge.”
Researchers also explained how it is more difficult to evaluate a move with the multi-colored puzzle box compared to that of Chess and Go’s which can be determined whether a move is closer or further to a favorable outcome.
McAleer also said that the method could also be applied in protein folding aside from solving Rubik’s Cube puzzles.
“We are working on extending this method to find approximate solutions to other combinatorial optimization problems such as prediction of protein tertiary structure,” he said.
(Photo source: universita.it/ techspot.com)