Assuming you mean you’re writing more concrete (less abstract) code as a means of productivity, if you want to go back to writing abstractions, you might need t

Assuming you mean you’re writing more concrete (less abstract) code as a means of productivity, if you want to go back to writing abstractions, you might need to go into academia. That seems to be the only place where the necessary years of thought go into creating abstractions that actually stand up to the problems they intend to solve. I’d cite Haskell and GraphiViz as examples of this. Personally I’m also interested in deductive databases and ontology but, on the whole, those are too ambitious even for academia these days, which I think is a shame because to me logically organizing information is probably one of the most important problems in IT since the web was created.