The Development of Autonomous Virtual Agents

Abstract

This chapter presents a potential application area for SmartData (Tomko GJ, Kwan H, Borrett D. SmartData: The need, the goal, the challenge. Report, University of Toronto, Identify Privacy & Security Institute, 2012) research of importance in the near future. Technological and sociopolitical trends augur for development and adoption of a mobile distributed system for personal data collection and storage that incorporates the ideals of Privacy by Design (Tomko GJ, Kwan H, Borrett D. SmartData: The need, the goal, the challenge. Report, University of Toronto, Identify Privacy & Security Institute, 2012). Such a system will necessarily encompass a comprehensive interface which implements a complex data privacy, security, and sharing policy. This privacy management and sharing policy for distributed sensing participants represents a potential early embodiment for SmartData agents with unprecedented importance. Furthermore, distributed systems such as these form a convenient population of individuals embedded within the environment in order to exploit parallelism for crowd-sourced distributed learning. Such populations of participating users and their devices represent an intriguing opportunity to collaboratively develop a test-bed for the training and validation of SmartData agents directly within the target environment. Such embodiment and embeddedness within the 3D environment of the real-world forming a ``mobile cloud'' of Pervasive Internet devices is complementary to and converges with the vision of SmartData agents operating in virtual 3D online environments. A possible simulation test-bed for gaining insight into evolutionary dynamics in such a distributed learning context is discussed.