The word "Abstract" is used in many contradictory ways, but it's usually gesturing at the idea of "distance from reality".
In computer science, there's an idea of "higher-" and "lower-level" programming languages.
Higher-level languages are more "abstract", they say, because they organize machine code using very specific patterns, and then force the programmer
to work only with the patterns, forgetting that machine code even exists.
This is, of course, secretly getting at the broad idea of emergence.
The universe is Quantum Field Theory, deep down, but if you throw a rock, you don't need a PhD in theoretical physics to predict where it'll land.
That's because we have Classical/Newtonian Mechanics, a simpler, higher-level, "abstract" model which takes advantage of common organizational patterns in Quantum behavior.
I find it interesting that Computer Science tends to be about "built-in emergence", designing systems intentionally such that a useful higher-level model is available.
But, wait, Quantum Mechanics is the abstract one, not Newtonian Mechanics... Right?
I suppose what people normally mean by "abstract" is something more like "distant from my everyday experience".
After all, we don't actually know what the "final theory" of the universe is.
All we really know is higher-level organizational patterns.