A group of artificial intelligence researchers have created a version of the Super Mario World game where everyone’s favourite Italian plumber has become self-aware.
The team from the University of Tubingen have programmed Mario to respond vocally to questions and commands, while he can also act upon feelings of hunger and curiosity.
This Mario detects hunger when he hasn’t collected coins in a while and autonomously goes in search for them, effectively meaning he begins playing the game himself.
Thanks to the integration of Carnegie Mellon’s speak recognition toolkit, Mario can understand spoken commands, meaning he can act upon the information he’s given and respond with natural speech.
So, for example, once he is informed that jumping on a Goombah will kill it, he can then do that for himself. He can discover that knowledge for himself if he is told to search for an enemy.
“We took a Mario clone that has been developed by others and then we gave the agent basic knowledge of what his behaviour does,” says Martin Butz, head of cognitive modelling at the University of Tubingen told Newsweek.
“We have equipped Mario with an internal motivated state, for example to collect sufficient coins whilst he is interacting with the game. We give him internal needs – what we call a constant homeostatic state – like hunger, and whenever this equilibrium becomes unbalanced Mario learns to respond based on his previous interactions with objects,” he added.
You can check out the fruits of the research in the YouTube clip below.