Hacker News Reader

Try playing an adversarial word game with ChatGPT. like the rules are, one player asks questions and the other is not allowed to say "yes" or "no", not allowed to reuse the same wording, and not allowed to evade the question. You'll see its tail wagging pretty quickly.

View on HN

You could very likely train an AI (llama?) to easily do this but trying to get a general LLM to play a game such as this doesn't make sense. Best way to get around it? Have it create a python program that will play the game correctly instead.

Reply on news.ycombinator.com