"What you gotta understand is that in the beggining of Electronics, people were pretty much trying to put two materials together they thought worked and then tried modelling it. It was pretty much trial and error, experimentation..." and then he hits me with the most "holy s*" moment of my academic life: "... much like programming and software engineering is today. You write some code, run it, see if it works. Works, ok, go ahead, make sense of it, explain in the documentation, next task".
I had NEVER thought of software like this, it just hit me like an atomic bomb in the head, I felt like I understood where in the history of software engineering we are right now. Structure is coming, slowly but surely.