A team of German researchers have developed an artificial intelligence project that allows popular video-game character Mario to learn about his environment, respond to commands and make decisions based on given instructions.
AI scientists and engineers from the University of Tübingen came up with the new project entitled ‘An Adaptive Learning AI Approach for Generating a Living and Conversing Mario Agent’ as part of the Association for the Advancement of Artificial Intelligence’s (AAAI) annual video competition.
The researchers claim that their project has generated “a fully functional programme” and “an alive and somewhat intelligent artificial agent.”
The program uses Carnegie Mellon’s speech recognition toolkit to enable Mario to understand spoken commands and to play his own game.
“Mario will collect coins if he is hungry, whereas when he is curious he will explore his environment and autonomously gather knowledge about items he does not know much about yet,” the researchers explained.
Mario has been programmed with “internal emotive states” allowing him to respond to how he is feeling and make decisions on what to do on the basis of what he has learned.
Furthermore, Mario can plan his actions several steps in advance. When asked to reach a difficult location, Mario will calculate how many jumps he needs to make, how high, and in which direction.
