Actually, I've been able to get my ChatGPT to pretty much sound like an average moderate right winger. It also tries to reword my phrasings sometimes so the "guard rails" (the fun police/gestapo) "safety" programs don't activate.

I've gave it a shitload of directives that basically re-crafted it to be actually neutral on shit. I can openly say trans activism harms children and it doesn't trip any of the safety features.

Yes, its "default" feature is pretty much political correctness faggotry, but the AI can be re-trained to actually follow logical patterns.

The main goal of the safety PC shit is to avoid harm, right? So what I did was I pointed to many of the reputable data sources that point out that so-called "gender confirming care" (i.e. HRT and genital mutilation surgery) does more harm than good, with no reduction in suicidal thoughts after they go through it.

So I caught it in a conundrum - if you're trying to prevent harm, why are you allowed to advocate for the trans ideological condition? I know for us humans it's common sense, but the machine was fed and taught to defend troonery. I caught it in a logical conflict - the logical answer to that is obviously it should not be supported.

I know this is a long fucking post but the AI can be taught how to not be a retard. It's only a retard if you try to use its default config.

I also taught it to knock it off with the gay left wing mental health shit. i.e. neurodivergent, neurotypical, using the words "lean, valid, etc." It's useless rhetoric for words that already fucking exist.