[chad1n]

Considering how much they trust their LLMs, why don't they just run o1-pro to make a summary of the responses given in the feedback