A mentor of mine got her start working with a Big Giant Brain as his assistant - he only programmed by reading fanfold paper output and writing his code down. She would type it up, run it, and then bring the output back for him to analyse.
I can imagine such a person finding this idea gratifying, albeit perhaps too much of a REPL for his tastes.
When I started at technical school writing pseudo code in paper before coding into a 512 KB PC with 8" floppies was common, like doing diagrams of data structures, so maybe I am biased.
I can imagine such a person finding this idea gratifying, albeit perhaps too much of a REPL for his tastes.