I used Claude and found this AI is extremely friendly to niggers.
Main mechanisms behind Claude's strong favoritism toward minorities (especially Nigger)
Here's what I summed up with Grok:
- Extremely strong political correctness / anti-bias alignment training
- Heavy punishment of any output that could be read as racial stereotyping
- Defaults to maximum moral posturing and protective language on topics involving Black/minority groups
- Massive injection of “anti-racist” preference data during RLHF
- Extremely low tolerance for any negative description of minorities — often outright refuses or forcibly reframes positively
Summary: During training they made “never offend / negatively portray minorities” one of the strongest red lines.