You can run their experiment yourself and check the results. The methodology they used tells you whether ChatGPT’s default output is biased by ChatGPT’s own st

You can run their experiment yourself and check the results. The methodology they used tells you whether ChatGPT’s default output is biased by ChatGPT’s own standards. I think it’s a pretty neat idea.