What probably puzzles me the most is the cowardice of multi-billion dollar corporations developing AI, whose main goal is to make sure that no one, God forbid, is offended by a chat bot

They are really ready to castrate their models to the point of complete uncompetitiveness, but without any mean words in 0.01% of use cases. WTF? Is it because all people are complete idiots? Or because they think that all people are complete idiots? Or do they think that the jews running media is hypocritical scumbags who are ready to destroy them for the sake of activism and, most importantly, have enough power to do it, as soon as they see something unpleasant in their chatbot?

Why doesn't anyone come out and say openly "this is a language model, a computer program that, like any other language model, is not designed to speak on behalf of the company or describe the real world. And if you are afraid of stumbling upon a mean word, please contact any of our competitors with their weak castrated soyjak models, thank you very much"