The technique can be interpreted in different ways. One thing that definitely helps is when there’s a strong hype cycle, like with AI, where you’re tempted to build the next “super framework” only to realize that another team may develop it faster and better than you ever could.

I have a friend who worked at an AI company before the current boom, and he once told me something along the lines of: we built several things over the past few years that could now be replaced by the new frameworks that keep appearing.