Why many important minds have subscribed to the existential risk of AI

Elon Musk has been sounding the alarms about the existential risk of the human species due to Artificial Intelligence. He's a brilliant leader and has one of the sharpest minds in the public eye, so for him to be alarmed about a field I work in and sense no such existential risk has caused considerable cognitive dissonance. My instinctive disagreement with someone of his stature, along with people like Nick Bostrom, Stephen Hawkins, and Bill Gates made me curious. Nick Bostrom wrote the best seller Super Intelligence that describes a world where machine intelligence dominates human intelligence and ends in our extinction. As bleak as this world view is, Bostrom approaches it with reason and with loose timelines. He makes a strong argument for working on the AI Control problem, in which we humans have the ability to easily shut down an AI if it misbehaves.