A Harvard University student suggests that reading more chatbot-generated content reveals a 'peculiar superficiality' in AI's prose.