Developing a macro cognitive common model test bed for real world expertise.

Abstract A central challenge for the common model and for Artificial General Intelligence (AGI) is the development of functional human-like agents capable of performing complex tasks in dynamic real-world environments. The cognitive test-bed model approach is a method of applying Allen Newell's Single Complex Task Analysis method to create a human-like AGI. In addition to the environment, we include expertise in the test-bed. The expertise is placed in the agent, and the agent is placed within the environment and provided with complex tasks to perform. The agent development process and agent task performance was gauged for backwards compatibility in the performance of open-ended tasks. [ABSTRACT FROM AUTHOR]