One of my random theories of engineering is that as we’ve gotten better and more precise in our computer modeling we’ve also gotten overconfident in our own modeling abilities.

So we no longer over design with random “engineering factors” added in for the unknown, saving on costs. This of course falls apart when reality differs in some unexpected way from our modeling and it turns out our “just cost optimally regulation safe enough” design actually “just shy of being actually safe”.

The 737max debacles seem like patches on patches on patches trying to thread that needle.