Strategies for Helping SISSY Systems Deal with Knowledge Gaps and Unknowns

The purpose of this position paper is to present first what may appear as a philosophical problem - the limitations of knowledge by any single entity - and then show that key strategies in biological systems for somewhat overcoming these limitations lead to a pragmatic emphasis on the development of certain capabilities in SISSY systems. We first describe three types of considerations that we and intelligent animals do in regards to unknown environments; these include ways to gain more knowledge about an unknown event or object, ways to behave in the face of different degrees of confidence that one can still do a task given the unknowns, and lastly, the need for adaptive mechanisms that allow us to respond rapidly and as safely as possible to new events in the environment. These rapid responses are both for suddenly occurring events or objects that may harm the intelligent system and also for taking rapid advantage of sudden opportunities that may benefit the intelligent system. We then discuss the implications for SISSY implementations of each of these types of considerations.