AI risks are something humanity can manage - let's get working on how ...