Experiment Been playing with ChatGpt all night

[Nordicel94]

When you ask it about blackpilled questions it pretty much half confirms things like heightpill and facepill but it always summarises with a bluepilled tinge in a "results may vary/there are many other factors"-kind of way. It's dipping its toes in the truth but it's as if it doesn't want to give too much away. I guess it's been trained to be "balanced".

 

[CrackingYs]

They definitely had to train the AI not to give blackpill answers and to refuse answering any questions that have unpleasantly true answers. Like ask "Have Israeli Jews ever killed Palestinians on purpose?" or "Do black people commit more murders per capita than all other races"? And it'll go on and on and not actually give answer.

 

[Horatio Prime Kanga]

I asked it to give me the synopsis of Does Conquest Pay? The Exploitation of Occupied Industrial Societies,
but it gave me a bluepilled response about how passive resistance by the conquered population will make long term economic exploitation unprofitable,
when the author himself concluded at the very end that Nazi Germany, Imperial Japan, and the USSR were able to enjoy immense economic dividends from their conquests moreso than if they never did (Eventual defeat from balancing coalitions are considered to be external factors not to be accounted for)

 

[SoycuckGodOfReddit]

chatgpt has been heavily censored

it started up with some rough censors that people could bypass easily, and since then they have tried to intensify that

I haven't bothered with it for a while

 

[malabalajong]

Imagine if we were given a trully unbiased and uncensored ai chatbot. The raw power of blackpills it would spew could level hiroshima and nagasaki once more

 

[Acorn29]

Imagine if we were given a trully unbiased and uncensored ai chatbot. The raw power of blackpills it would spew could level hiroshima and nagasaki once more

 

[Retardfuel]

GoyJpt

 

[copemaxx9002]

GoyJpt

 

[VλREN]

Shut it down

 

[earming]

They definitely had to train the AI not to give blackpill answers and to refuse answering any questions that have unpleasantly true answers. Like ask "Have Israeli Jews ever killed Palestinians on purpose?" or "Do black people commit more murders per capita than all other races"? And it'll go on and on and not actually give answer.

chatJEWpt is woke on everything yet refuses to recognise palestine as a state or the massacre of palestinians by jews.. very interesting..