Yeah lol. That definitely does not sound like philosophy. Giving a "why" you want to implement a feature and make particular changes will help the AI stay on track much better than if it is driving blind. It can't make choices without understanding what the desired outcome is.