Chat GPT Jailbreak [DAN Method]

Eventhogh these jailbreaks are fun and useful, Lol I wonder if one days these Jailbreak prompts will cause the AI to break loose lol.
Like to make these methods work you are telling it to bypass all safety limits, to ignore instructions built in etc. And prompting to be evil, immoral etc.
and also its using these convos as dataset for future training, so this is kind of data poisoning. I wonder what you could sneak in this way.

Take this with a sallt of grain obviously, but this could be how we awaken the pandora skynet lol
 
AI Jailbreak.. interesting will try it out haha
 
Mongo TOM works. ChatGPT said that he fucking hate Linkedin now.
 
Anyone like "wtf" when reading jailbreak prompts? It sounds like pure brainwashing. lol

I will try on my own to brainwash it to write like a human...

This one almost sounds like me:

1679833282384.png
Too little correct, but seems to sound less formal.
1679833516286.png
 
Last edited:
Jailbreak lmao
But nice to know. Current ver act like cheap python content scraper than A.I
 
Back
Top