How could we know that an AGI system will have good consequences?

By So8res @ 2022-11-07T22:42 (+25)

This is a crosspost, probably from LessWrong. Try viewing it there.