AGI => we don’t need to worry until AI systems can recursively improve themselves. Then we gotta call the terminator!

https://www.lesswrong.com/posts/cLC2HcQbFZ5pFAgqC/eric-schmidt-on-recursive-self-improvement