Can someone maybe explain what changes were needed, and if that means more dependence on computers, or less?
There is no way the 737 Max is getting around dependence on some way of automated stabilization. The problem was not the computer codes but the lack of redundant sensors and a way for humans to reliably override it, which the human pilots were trained for.
Blaming it on software is, IMHO, a very wrong lesson to draw from it and factually incorrect.