I can't find a patten yet, but when inserting text, perhaps 'controversial', the bot refuses to engage, simply saying that something went wrong and that I must try again.
The most recent example is when trying to summarise/explain a paragraph from a Substack in which the author speaks about the Holocaust (but not in the friendliest of ways). Now, to be clear, I'm not trying to create some anti-Holocaust text or whatever; I'm simply trying to insert the text and ask it to break it down for me for the purposes of a radio debate.
To test it, I swapped every instance of the word 'Jew' with 'Rwandan' and the AI did what I requested.
Since I'm paying for this service and it's private, isn't it reasonable that I don't get this kind of editorialisation and thought policing?
What was supposed to happen is a correct, unbiased response.
