When we encounter a fresh object,scene it’s useful to make judgments about which parts of the scene are a) stable & literally encounted before, c) stable & closely similar to elements from our past experience, d) novel, but expected to follow known prototypes in certain dimensions, e) totally unfamiliar & hard to predict.

Intuitively, the system that mirrors human levels & types of intelligence has to store enough detail to reliably make those sorts of judgments.

Real world situations come about as a result of real world historical processes, both immediate & near & long term & evolutionary. Human intelligence makes use of historical process coherence & story in evaluating scenes & making inferences about them. Artificial reasoning schemes that use situations & VR will miss some of the coherence of human conception if they do not include constraints for plausible/likely historical evolution that led to the given scene interpretation.