First, AI won't call for human-like "sentience" to generally be an existential hazard. Modern AI plans are offered specific targets and use learning and intelligence to obtain them. Philosopher Nick Bostrom argued that if one particular gives Virtually any target to a sufficiently high