Ask HN: What do you think of this prompt as a benchmark for LLMs?