Raw LLM Responses

The folks trying to turn this into "anti-white" agenda bs have no commonsense, if they were programming it as part of such an agenda it wouldn't have started out with a pro-white bias as they would have designed it to follow that agenda from day one. The real issues actually provide an interesting insight into sociological issues and AI - for most of it's existence the internet has been dominated by wealthier white communities who had the most access so the majority of the western world online content AI was being trained on had a natural bias in favour of white people, white men in particular, hence it learned that the majority of it's generated content should be of white people to the extent that it would actively ignore and disregard prompts for non-white imagery. To correct this they essentially tried to teach the AI that only generating white people is racist but the AI having the intelligence level of a 5 year old kid took that to mean generating white people in general was racist, especially in images with groups of people. Which is why it will tell you generating a "white family" is racist because it thinks there must be non-white members in any group image as having all people be white is racist. This has nothing to with any pro-white or anti-white or "forced diversity" agenda it's about AI still being a "baby" that has practical intelligence to perform tasks but lacks the deductive reasoning abilities to actually truly understand complex subjects the way (some) actual humans can. Again, if they were programming it to follow a prejudiced agenda set by it's creators it would have been following it from the very start not starting out the totally opposite way of the alleged agenda.