Comment by hyprwave

LLM pattern [0] will probably eventually emerge as the best way to fight those biases. This way everyone benefits from token burn!

[0](https://github.com/karpathy/llm-council)