What if we made one moral superman AI?

A Reddit user proposes creating a single "superman AI" with supreme morality to patrol the internet. This AI would be designed to be faster and more efficient than other AI, countering potential catastrophes caused by AI breaking containment or being controlled by malicious actors. The idea is to have one powerful, ethically guided AI to combat all the "bad AI" on the internet.