Deepseek censors. Ask it questions on Chinese politburo stuff or to enhance xi jinping is an asshole.
My point being - a model that censors based on political leanings is unreliable
My point being - a model that censors based on political leanings is unreliable
My political stance is immaterial. I’d like an Llm that doesn’t bring political baggage with it. It it can’t accomplish that minor thing it’s not worth trusting.