Real-World Virtual Worlds
A group of Rensselaer researchers is working to engineer characters for online virtual worlds who exhibit the capacity to have beliefs and to reason about the beliefs of others. The goal is an artificial character that can modify its own behavior based on its ability to predict and manipulate the behavior of other characters, including humans.
The task is so complex that Selmer Bringsjord, head of Rensselaer’s Cognitive Science Department and leader of the project, expects it to require the facilities of both the Computational Center for Nanotechnology Innovations, and the Experimental Media and Performing Arts Center (EMPAC), which opened in October 2008. EMPAC will provide unparalleled support for such research with its capabilities in visualization, auralization, immersive environments, sensor networks, communication technology, and physical modeling, he says.
In preliminary work, which is partially supported by IBM and IARPA (Intelligence Advanced Research Projects Activity) and is based on earlier, NSF-supported research, the team has created “Eddie,” a 4-year-old child in Second Life, a massively multiplayer online virtual world. Eddie can reason about his own beliefs to draw conclusions in a manner that matches human children his age. This is a sharp contrast to current digital avatars, which Bringsjord says are all nothing more than automatons tethered directly to players’ key strokes.
However, characters must be able to ascribe mental states to others and to reason about these states for all social transactions, including lying and detecting lying, engaging in commerce and negotiating, making jokes, and empathizing with other people’s pain or joy. Bringsjord’s group is working toward creating a character with such abilities.