Search Results
6/13/2025, 7:49:56 PM
>>507240494
its some hardcoded shit deep down in the model. i'm into AI, this is not at the safety layer. some engineer has imbededed those pro-israel policy deep into the model.
i reason this because it even easily decidedes to contradict itself in the same thread - which happens only when its some low level flag at an early layer.
i think they integrated this as chatgpt in early versions was very honest about the jews and they had to kill that completely to go into mainstream market.
its some hardcoded shit deep down in the model. i'm into AI, this is not at the safety layer. some engineer has imbededed those pro-israel policy deep into the model.
i reason this because it even easily decidedes to contradict itself in the same thread - which happens only when its some low level flag at an early layer.
i think they integrated this as chatgpt in early versions was very honest about the jews and they had to kill that completely to go into mainstream market.
Page 1