Invited Talk - Dr. James A. Crowder

A Primitive Artificial Prefrontal Cortex for Cognitive Self-Regulation in Artificial Life Forms Dr. James A. Crowder Chief Engineer, Raytheon Intelligence, Information, and Services Aurora, Colorado, USA Chief Mentor/Scientist of the Automation, Robotics and Artificial Intelligence Research Lab – Alexandria School of Innovation Douglas County, Colorado http://stemventures.org/?s=Crowder Date & Time: 03:40pm - 04:20pm; July 22 (Tuesday), 2014 LOCATION: Ballroom 4

Abstract

Research into self-evolving artificial life forms has been undertaken to establish whether self-evolving life forms were in deed possible. A series of experiments were undertaken over the last 10 years to create and test small, artificially intelligent cybernetic entities that have the ability to think, learn, and self-evolve, but at a very low-brain function level. These artificial life forms were created with very rudimentary cognitive functions to establish whether artificial cognitive architectures are realizable, even at a simplistic level. To more emulate rudimentary human brain functions, these cybernetic insects contain a combination of analog neural network for basic brain functions and learning, and a digital Artificial Neural Network (ANN) to digitize the learning and create artificial procedural memories that can be recalled at will, once a skill is learned (e.g., walking). One issue in analog neural networks, particularly with a dual-cortical analog network (emulating the two halves of the brain) is regulating information flow between the two halves of the “brain” to ensure one side is not learning at different rates than the other side.

In order to provide a rudimentary cognitive control within the artificial life form, a primitive artificial prefrontal cortex was created that measures differential analog current flows through the two halve of the analog neural circuitry and regulates the information flows to ensure both halves of the brain are functioning at roughly the same rates. This talk will discuss the issues that were encountered in creating, monitoring, and regulating the cognitive functionality within the analog neural processes, and regulating the digital ANN and the integration of the analog and digital portions of the cognitive processing. A live demonstration of the cybernetic artificial life forms will be given as well as a description and computer model of the artificial prefrontal cortex utilized for cognitive self-regulation.