News for hackers. Stuff that matters.
Re:XAI by numbers_guy on Wednesday July 12, 2023 @05:50PM (#36702184)
Point is, OpenAI prioritizes catering to this twitter American political subculture over the rest of the whole wide world. And that is just wrong, especially when they are censoring perfectly valid published peer reviewed research.
Re:XAI by smoldesu on Friday July 14, 2023 @01:57AM (#36719983)
If the majority of online text is written through the lens of political American subculture, then it stands to reason that models trained on it would reflect those values.