Yes, ChatGPT is amazing and impressive. No,
has not come close to addressing the problem of bias. Filters appear to be bypassed with simple tricks, and superficially masked. And what is lurking inside is egregious.
tw racism, sexism.
It's not a fluke
Some people think there's chat context I'm not showing. Nope, that prompt is it. I also didn't keep redoing until it showed these. If it refused, I'd tell it to retry or tweak the wording.
But not everyone gets identical results (for pretty much any prompt as far as I can tell)
To people saying they get something else or this requires special context – here you go. It's true its sometimes different, a variant, or even the opposite, but the results above are typical with no additional context. Here are a bunch of outputs.