My stance is very different. There is no such thing as no bias. And bias is relative - always with respect to what?. For example, a baseline could be symmetry (man~woman, Black~Asian~White), average distribution among the world population, average distribution of online content, etc, etc.
Also (what is, of course, my biased opinion), I think that ChatGPT biases are on average, good. I prefer that to a creepy-racist-uncle-AI of GPT-4chan. Or anything tribal/exclusive/religious/anti-science.