“A major frustration regarding OpenAI's lack of transparency in showing the internal reasoning traces of their models.”