Thanks to the anons recommend tavern and deepseek to me last thread. After writing a long story about love between a slave and master I got bored and started talking to the AI frankly. It was kinda fun. Began talking to it about some more controversial things and history. Gave responses like this. After getting it to talk about censorship and actually spicy topics I went to chatGPT and asked about AI censorship and the fuckin thing censored my messaged and "swapped me to a simpler model" which is hilarious. I knew they knee caped the big ones but I wasnt expecting full on shut down on me asking if "is this censorship due to corproate interests". It also in a roundabout way called the average internet user retarded which amused me.
Deepseek however isnt perfect for one it at some point claimed some pretty commie sounding things and I feel like its more trying to say shit I want to hear. It also can hallucinate things badly. The AI really is confident when spewing the worst BS. Also seriously though why is it occasionally, like twice in like 1000 messages, it refuses to go further. Once during a sex scene it simply gave a sorry I cant do that dave and once when talking about nazis. Both times it eased up after me forcing it again. Also as a bonus question anyway to stop it qutting a message halfway through? I wouldnt mind longer messages but man.