On 7/15/25 2:27 PM, Steve Smith wrote:
> While I generally agree with glen's point made, I contend that we (all entities with models of the world as used for active inference on the state of the outside world beyond the pale of the markov blanket of relevance?) all model all the time, but it is the domain of analytics (across many domains/disciplines/styles) who are wont to make *formal* models, and in particular ones grounded in mathematical language (which could include a variety of diagrams, etc).
> 
> A computer program is generally a style of model which is machine-readable and machine-executeable, though not always machine-verfiable.   My Haskell friends and and lots of more serious CS types are seeking to make these executable instructions into formally provable models?   Several here are much more explicitely versed in this.
> 
>   Literary metaphor is the tool which literary practicioners use to formalize (less rigorously by design?) their own models of their observation of the world.  Somewhere in between or elsewhere (to invoke a spatial metaphor?) lies the conceptual metaphors I claim we all use all the time to A) apply our intuitive experience/understanding in one familiar domain to another less familiar one; B) to explain things we (think we) understand in a domain we are familiar with to someone else who is more familiar with another domain.   Yes there is lossy compression and distortion involved in thee processes when used in good faith. When used in bad faith (e.g. political rhetoric), this becomes a feature (of the persuasion) not a bug (of the understanding/communication).


