All applications are just "inventories" of arbitrary data.
A while back, after years of software design and development, I've come to realize that most of the software out there, is in it's essence; an inventory data tracking system.
It all comes down to state machines and the most efficient way to funnel the specific type of data, generally coming from the most obscure corners of a hard-coded file, to the live rendering of the last just-typed key code by the user. All you need to do, is make the data travel throughout the system as seamlessly as possible, to the end of the current data-layer.
While I do think its important to make distinctions between an app specific goals and objectives and its architecture, it is my opinion that following "CRUD" practices makes us follow the path of "all apps are just state machines".
Chatting with LLMs about it, it is clear that they rather validate "all apps are just state machines" over "all apps are just inventory-like systems" because it seems that generalizing is frowned upon among LLMs, but ultimately they do agree on the general statement and the abstraction of the core idea, that most apps are doing, is receiving, computing and giving data to the user, meaning the magic is indeed gone from all conventional apps.
LLMs however give a glimpse of hope into the future of software architecture and the mystical ways "applications" or "higher-intelligence" entities evolve. Unanswered questions for the future, what is software, what is an LLM? what is an entity, and again; what are applications?
Top comments (0)