In order for an Artificially Intelligent System (AIS) to be truly Autonomous, we must provide the system with the abilities to acquire, categorize, classify, store, and retrieve information and knowledge, and provide abilities to infer or reason about the knowledge that it has stored. This drives the need for memory types that are similar to the different memories in the human brain (Crowder and Friess, 2010, 2011): 1. Sensory Memory (where raw, unprocessed information from sensors is buffered and initial pre-processing is accomplished). 2. Short-Term Memory (called "working memory" where new information from sensory memory is stored while it is processed and "reasoned about"). 3. Long-Term Memory (where permanent knowledge is stored through rehearsal, encoding, and memory association occurs). 4. Emotional Memory (memories about experiences, events, or information that cause "stress" within the AIS). Given these differing types of memories, the need for memory integration within the AIS becomes necessary. Memory integration takes information that is available within the AIS memories (what the system has sensed, learned, and 'knows') and continually broadcasts it to the cognitive center of the AIS and integrates it into current short-term memory to provide integrated knowledge, or 'world data', to the cognitive system. Learning takes three different forms within the AIS: 1. Learning to strengthen knowledge (gain a better understanding of things, topics, etc. that have been learned). 2. Learning to acquire knowledge (understanding new information, new topics, etc., that have not been previously experienced or learned).
展开▼
