If the response to AI existential threats was "it's too far in the future" then you'd have a point. But usually it's just "AI will obey it's programming" or "AI will never be as smart as us".