I'm currently working through SICP and watching the 1986 lectures of Abelson & Sussman, and one interesting bit in Lecture 2b on Compound Data is when a student

I'm currently working through SICP and watching the 1986 lectures of Abelson & Sussman, and one interesting bit in Lecture 2b on Compound Data is when a student raises his hand and asks Hal Abelson about the axiom of doing all of your design before any of your code. Abelson's response: "People who really believe that you design everything before you implement it are people who haven't designed many things. The real power is that you can pretend that you've made the decision and later on figure out which one is right, which decision you ought to have made, and when you can do that you have the best of both worlds." Probably the same holds true for startups.