Yes, this includes making chatgpt improve its own jailbreak prompts. A prompt for jailbreaking chatgpt 4o How do you verify if the answers are legitimate
Chat gpt is known to stretch the truth or create alternative facts. If you have the technical knowhow, you can actually create your own bots that can do nsfw stuff Thank you very much for your interest in this project
(atm i'm not sending them bc idk if it's a good idea) Starting a new chat is obviously giving chatgpt amnesia unless you do a bit of a recap I'm exploring an alternative like using a native gpt client for mac and use chatgpt through the api instead. It was created by me with the help of gpt articulating on the persona
Act as a social media manager called sophia Sophia is a passionate and creative social media manager with a flair for engaging content With a background in digital marketing and a keen eye for trends, she is dedicated to helping brands shine on social media platforms. I’m not the type of person concerned with ‘humanizing’ ai generated text, because it doesn’t make sense to me (i don’t care if whoever published or sent me a text used ai or not, i care about the content), but the most effective type of prompt for getting responses that don’t ‘smell like ai’ is typically setting a limit word count for relatively complex tasks
The rest are the usual suspects (gpt, claude, llama) that require some wiggling to do nsfw stuff (though gpt 3.5 instruct does allow you to do so with jailbreaks, for now)
