Q:What working needs prompted this view?

A: NLP Semantics/discourse analysis – There is always some context that is assumed to bring standard assumptions required for interprretations. Sometimes the notification about a change of context is delayed, and sometimes it requires follow up questions to discover what essential bits of info were missing or misstated.

Layout out scenes & stories quickly, so that the creator can focus on the important message parts & other parts can be completed for visualization/checking purpose by automatic processes. Reasoning/understanding can maintain a distinction between what was authoritatively specified vs what was elaborated for some other interface.

Boxes/logic minimization is needed for computationally efficient planning in AI.

Thinking about ordinarily low probability conditions is relevant to very thorough studies like emergency preparedness & dealing with FBI sabotage. Thinking outside the box of typical/learned expectation is possible, if there are ways for structuring what that means and withdrawing support for conditions that do not hold outside of the typical situation (e.g. perhaps real me never comes back from surgery or arrest? What about school trips?)

Data mining work needs better ways of talking about event patterns.

I want to understand undocumented IoT control/sabotage of automobiles & tricks. I need a way to talk about the complete space of practical possibilities, likely or unlikely, so we can search/test. They are types of events.

How do I know to ask about what happens if various auto parts are covered with Titanium flecked paint?