How do silly videos align with OpenAI’s goal of building ethical artificial general intelligence for the good of humanity?