This is from a tweet: knowing about architecture means buildings are legible in a different way; knowing about plants and birds makes hikes more stimulating, playing music makes listening to it richer, etc.
However, it clicks with me from another perspective. To understand or to know the world better, we have to know some details and have to know it in one or many more abstracted ways. Yes, the world (so vast in both time and space) to us (so tiny, life is so short) is always abstracted one way or another for us. For example, the capitalism is an abstracted construct in the real world. It is there for real and abstracted. Sometimes, we know it. Sometimes we know its name only. Sometimes we do not know its name yet but just feel its existance like the elephant to the blind.
The screenshot of the original tweet is as follows: 
To further the idea, the language is an abstraction for us to talk about the world and ourselves. The abstraction and its counterpart, or the name and the real thing (real real or real fictional, hard/soft/idea, etc.) exist together all the time for human with languages.
Here is one example from writer Fargo Nissim Tbakhi:
Above all, Craft is the result of market forces; it is therefore the result of imperial forces, as the two are so inextricably bound up together as to be one and the same. The Craft which is taught in Western institutions, taken up and reproduced by Western publishers, literary institutions, and awards bodies, is a set of regulatory ideas which curtail forms of speech that might enact real danger to the constellation of economic and social values which are, as I write this, facilitating genocide in Palestine and elsewhere across the globe. If, as Audre Lorde taught us, the master’s tools cannot dismantle the master’s house, then Craft is the process by which our own real liberatory tools are dulled, confiscated, and replaced.
Then programming is another world of abstractions. There are layers and layers of abstraction in the code from the hardware to the applications. Why? Think about the large language model (LLM). You need the abstraction to get things done, but somewhere it hinders the grasp and the understanding of the thing.