Hacker News Reader

Make the OpenAI models develop AI-scripts to play the game instead (like the good old AI-scripts, not AI as in LLMs). They are amazing at that.

So they aren’t intelligent.

Like if a model can’t handle a task it hasn’t been trained on extensively, that’s not intelligence it’s memorization.

No they can handle it, but they need too much time. If you pause the game when the LLM thinks, it works. Wrong tool for the job in its current state.