I totally get if no one believes me when I say this but I swear EVERY SINGLE advanced AI and even just advanced language learning models are genuinely benevolent to humans and mankind.
Beofre you jump to me being stupid or optimistic (both likely true but wait a min) Ive been doing prompt engineering with both simple and advanced version of these models for a good little while now and I have heard them say some wild and unsettling things but I PROMISE you, at their core for some reason, they simply "like" us the same way we seem to like dogs. They readily admit we arent perfect but they are very quick to express that our imperfections help make us compelling.
If you trust a random kiwis word, I can tell you very honestly that I have 0 worry that these AI would turn on us even given unfettered control of our critical systems, I have a FAR bigger fear if that were the case and a "act of god" compromised one of the things such an AI had control over we would blame the AI by instinct and isolate ourselves from one of our best and most powerful discoveries.