My name is Nick DePalma, and I am a Ph.D. Candidate in the MIT Media Lab working under Dr. Cynthia Breazeal. My passion is to push the bounds of what embodied artificial intelligence is capable of. I write software and design cognitive architectures for interaction. I’ve spent most of my life hacking (writing software) that is informed by reverse engineering devices and arguments. I do this just to understand what the designer was trying to do. What drives me these days is a dream of robots working with us in close proximity, building robots that are designed for us. Academically, it’s natural that I’m interested in social cognition and pragmatics and how observations and ideas from brother and sister domains should inform the design of the intelligent system embodied in the robot. Much of my vision is to integrate ambient and embodied social intelligence. Think the cross section of human-interested AI, like that which was embedded in fictional characters like HAL and C3PO.
I’m currently studying the behavioral effects of embodied interactions with a human from a pragmatics perspective, thinking deeply about the indeterminancy of referential gesture and it’s effects on learned representations. I also have a history in machine perception and entertainment. My dissertation contributions a visual attention system linked deeply with decision making regarding referential gesture and is designed to work on the humanoid robot Maddox. My dissertation also contributes a study of that architecture’s behavioral effects in an interaction. I make suggestions which improve participants engagement through the robots behavior.
ambient intelligence: Intelligence that is embedded in the environment rather than embodied in a movable device.
embodied intelligence: Intelligent machines that must deal with perception and action in a single, fully connected device.