I describe how binding safety standards including a red line for loss-of-control can lead to a future where society flourishes with guaranteed-safe tool AI.