Hacker News
Ask HN: How are Markov chains so different from tiny LLMs?