Ramana Kumar comments on [Link] Why I’m optimistic about OpenAI’s alignment approach