Two Way Language for Technology
10/16/2022 8:49 AM MST
language, programming
Designing a written/spoken language using the knowledge of computer software languages:
Forms – Platonist forms (like classes)
Qualities – An appraisal which grants a form an additional property
Concepts – Logical systems of qualities which may be applied to a form.
Logical Connectives – Semantic connectors which relate logic
Example:
I like comfortable chairs and dislike long commutes.
I – Form (self)
like – Function (applies the entailed semantics to the form (self) instance (I).
comfortable – Quality (modifies the entailed semantics of chairs)
chairs – Form (chair) and the “s” on chair infuses the Concept (plurality)
and – Logical Connective (both what came before and what will come after)
(I) – omitted, one assumes that the subject is the same if not mentioned
dislike – Quality (Subjective)
long – Quality
commutes – Form (commute) and Concept (plurality)
Structural logic:
I – Subject of statement
like comfortable chairs – First statement
and – separates statements while keeping them all true
dislike long commutes – Second statement
This structure could then be considered a Concept (perhaps the concept of divulging likes and dislikes.) As it is a concept, then it makes sense to assign it some word if it’s a unique one and may be used frequently enough. However, if we are referring to computers, there’s no real cost in granting it one.
Forms can obviously be part of a complex hierarchy. For example chair being of the form furniture, which is of the form of house placeables, which is of the form of human usable matter, etc. similar to inheritance in programming.
Qualities are like floating point variables since they can vary from 0 (moving away from neutrality in the negative direction) or 1 (moving away from neutrality in the positive direction.) If a single value is insufficient to describe the uniqueness of a form’s instance, then more qualities are needed, or perhaps a concept.
Concepts are sets of qualities and forms. Some qualities and forms may be predictive or more likely to be the case if an instance of a form or a form meets the definition of fitting a given concept. Qualities and forms may be required or not. However, in the case of a concept having many non-required qualities, perhaps a certain threshold needs to be met to be considerable as adhering to a given concept. Therefore, to describe a concept properly, one must be able to provide some formulation as to how to judge the relevant variables.
Qualities can be chained together. For example “An ornate comfortable chair” where both modifiers (one a concept and the other a quality) apply to the subject (chair.) However, there can be chains such as “A thinly boxed product” In this case, the quality “thinly” (the true quality being “thin” but modified in English) applies to the concept “boxed” but not to the form “product”. Perhaps this is because “thin” and “boxed” are compatible whereas “ornate” and “comfortable” are not. So in this case, knowledge of these qualities and concepts is required to understand how the logic applies. A safe bet in the case where one does not know is to simply apply everything to the closest form. Perhaps then, ornateness and comfortability are related somehow but this relationship isn’t known to the reader.
So far the system described only details what’s necessary (I hope) to describe the physicality and constraints of a logical system. However, it doesn’t “do anything” like this. Additional logical constructs are required in order to animate the logic. However, what exists hopefully can provide the foundations for a descriptive or evaluative language.
Thinking about AGI, I wonder how words are imbued with meaning. The most direct logical thing is to apply the definition to them, but that is simply more words. It must be some sort of cloud with probability or weights associated with the given word. This seems like the first necessary step to take when building one.
I wonder if constructing such a written language to command AGI and also to talk to others would be useful. We could build software much more easily, and any benefits or optimizations to the language we employ we could also use in conversation (perhaps with some explanation, although hopefully society will be quite intellectually enlightened in the future, by choice or by force.)