Academic
Research Statement
Everyone seems to be approaching cogsci from two distinct perspectives: top down (introspect, philosophy) and bottom up (biology, external observation). I think it’s fascinating to note that evolving artificial systems with the intent of deciphering their behavior and analyzing their structure bridges the two approaches. Given the ability to control the complexity of the experimental machine we can more easily observe both large-scale behavior and low level processes. Evolving intelligence in simulation gives the human experimenter god-like power over the developing agents. By placing ourselves in this position, we can control every aspect of the simulation whose analog in reality may be considered absurd. This intrigues me to no end, and is why I am spending so much time studying and performing research in this field.
Intel STS 2010 Participant
I participated in the 2010 Intel Science Talent Search with the attached research report on Integrated Information in evolved neural networks. I worked closely with Virgil Griffith (Caltech University) and Larry Yaeger (Indiana University). I employed the Polyworld neural network evolution environment and the Consciousness integrated information calculator.
Download PDF (500k)
Results from Integrated Information experiments. The data set includes anatomical neural network sampled from 5,202 Polyworldians, at death, from three genetically independent simulations. Each data point represents a 200 time step bin size. *First sample for which p < 0.01
