3 ms·

I wonder if it would be better to have 1 “perfect” LLM trying to solve problems or 5 intentionally biased LLM’s.