Evan R. Murphy comments on Reframing the AI Risk