“Existential” Risk Management for Artificial Intelligence (or just a little too much navel gazing?)