Saying "superhuman AI is not an existential risk" isn't the same as not caring about safety. It's a coherent assessment from someone working in the field that you may or may not agree with.