> But now engineering seems to be : let's ask my computing power what is the good design, with little further questioning.

I see this quite often at the undergrad level. Even for the simplest problem that can be solved with a pencil and a couple of calculations, many students launch their finite element software and blindly trust the results, even if it is telling them something obviously wrong, like orders of magnitude larger or smaller than expected, or with the opposite sign.

I acknowledge the problem, but I do not know where it comes from, or how to stop it.