Can the world’s first AI safety summit help avert the apocalypse?