> PoI asks these incredible questions: how do you design an AGI system that can optimize for saving lives, without it taking a strictly utilitarian view that leads inexorably to it trying to take over the world?

I don't think the show ever addressed those questions in a meaningful way? Basically there was a "good" AI, and later in the show, an "evil" one. And that was that.

The show did show us a convincing picture of the dangers of pervasive mass surveillance paired with an autonomous system that can sift through all of that data.