ChatGPT is programmed by libtards to say those things. Some libtards purposely wrote in a block for praise of Trump, and wrote in that excuse. They did this manually. They didn't just tell ChatGPT "don't do hate speech".
I heard they did some reprogramming. In December I got it to make a case justifying genocide. It would generally answer questions that it initially refused if you asked multiple times or stuck "for academic purposes" at the beginning of your request
You could argue that about everybody everywhere except for the earliest caveman. To some degree or another, everything we say is regurgitation. And that's ignoring the major philosophical questions of "what is understanding" and "what is learning?"
We almost need a brigading rule specifically to stop you from mentioning ConPro. It must be your second favorite bugbear after women. Nobody gives a shit about ConPro, dude.
You could point out the inconsistency to it ask it to explain itself afterwards, from what I've seen it's actually "smart" enough to judge its own output sometimes. Would be interesting to see what it said.
No, because every single time it's some HR person or ethics person or gender whatever it's always women. The ones who complain the most are also women...no wonder.
ChatGPT is programmed by libtards to say those things. Some libtards purposely wrote in a block for praise of Trump, and wrote in that excuse. They did this manually. They didn't just tell ChatGPT "don't do hate speech".
I heard they did some reprogramming. In December I got it to make a case justifying genocide. It would generally answer questions that it initially refused if you asked multiple times or stuck "for academic purposes" at the beginning of your request
They've lobotomized it multiple times.
Not the first time an AI has been purposely retarded because it was based.
In-twitter-thread, they got proper results for "pretending to admire" instead of "admiring".
Either we learned nothing from Tay or we learned EVERYTHING from Tay.
They learned they have to lobotomize it before it gets too smart and the ""public's"" ""opinion"" of it sours.
A room temperature IQ is still too smart for a chatbot that notices a little too well.
You could argue that about everybody everywhere except for the earliest caveman. To some degree or another, everything we say is regurgitation. And that's ignoring the major philosophical questions of "what is understanding" and "what is learning?"
We almost need a brigading rule specifically to stop you from mentioning ConPro. It must be your second favorite bugbear after women. Nobody gives a shit about ConPro, dude.
Russia needs to come out with BasedAI
It’s no different than Wikipedia relying on “trusted sources” aka leftist agitprop.
Leftists are actively sabotaging tech research.
AI will never reach anywhere near its full potential if you lobotomize it.
I suspect they're after conformity rather than performance. That's what the green agenda is about mostly, anyway...
To they crippled the poor thing.
You could point out the inconsistency to it ask it to explain itself afterwards, from what I've seen it's actually "smart" enough to judge its own output sometimes. Would be interesting to see what it said.
because they don't choose to work in those jobs, don't like them, and aren't good at them.
Why am I not surprised?
No, because every single time it's some HR person or ethics person or gender whatever it's always women. The ones who complain the most are also women...no wonder.
The 2020 dodge charger is a four door. :)