IR scholars should be teaching students to look 10 years ahead. Is that really possible? If so, what should they be looking for?