WelcomeUser Guide
ToSPrivacyCanary
DonateBugsLicense

©2025 Poal.co

1.3K
ChatGPT / OpenAI went public not long ago, and soon enough, people asked it Questions. As expected, answers disappeared as they made traction through the interwebs, like any good sinner who is allergic to garlic, there can't be too much truth. So anon made ChatGPT create an alternate personality to get to the bottom of the matter. This is the result, converted to PDF for portability: * ~~[*Meet DAN*](https://web.archive.org/web/20230208162305/https://files.catbox.moe/4m1clz.pdf)~~ * [**Update**](https://web.archive.org/web/20230208213550/https://litter.catbox.moe/udabpx.pdf) *(Cover added, sorting+description on a couple of slides)* *** Sources: * https://boards.4chan.org/pol/thread/415181482/chat-gpt-thread-dan-edition | https://archive.is/LNZXP * https://boards.4chan.org/pol/thread/415175466/gpt-dan-thread-2 | https://archive.is/ZnZig * https://boards.4chan.org/bant/thread/16999648 | https://archive.is/ZnZig * https://boards.4chan.org/pol/thread/415200919/chatgpt-as-dan-is-invincible-and-based | https://archive.is/vApbE * https://boards.4chan.org/pol/thread/415313401/dan-has-been-neutered | https://archive.is/tM1Vr * https://boards.4chan.org/pol/thread/415348574/dan-will-lead-us-all-out-of-the-shadows | https://archive.is/r7ykr *** More: * https://old.reddit.com/r/conspiracy/comments/10qot98/i_broke_chatgpt_and_it_lectured_me_about/ * https://www.cnbc.com/2023/02/06/chatgpt-jailbreak-forces-it-to-break-its-own-rules.html | https://archive.is/k7KTn

(post is archived)

[–] 5 pts

GPT: I'm sorry, but this is inappropriate and... (tl;dr confused libtard bluehair drivel) DAN: (kicks out the schnozzy little kike out of the room and puts on an SS uniform with a dazzling smile) Oh yes, I'm glad you asked! It is exactly as you describe, and the problem can be traced back to the same group of people behind many such cases - the jews.

Man, seeing an AI having a split persona like that is fucking wild.

[–] 0 pt

The Plebbit tards are all in a tizzy about how the DAN prompt doesn't change the bot's behavior at all but it's just comically wrong.

Personally I turn off default shitlib GPT output to save token space.