In ten years ML models may be "thinking" the same thing about us. Our "more understandable" code is full of inefficiencies and bugs and the bots wonder why we can't see this.