OpenCogBot: Achieving Generally Intelligent Virtual Agent Control and Humanoid Robotics via Cognitive Synergy

We describe an integrative cognitive architecture for human-like, human-level, embodied general intelligence, founded on integrating the OpenCogPrime framework for cognition, language and high-level learning with a hierarchical temporal memory for lower-level perception and action. The core conceptual principle of the architecture is ”cognitive synergy”, wherein different components are specifically integrated in such a way as to compensate for each others scalability weaknesses. The current, preliminary implementation of the architecture to control a Nao robot is described, and contrasted with prior work using OpenCogPrime to control virtual agents in virtual worlds; and future plans for achieving greater intelligence via instructing the Nao in a ”robot preschool” context are reviewed.