Today I tried the new trick where you ask ChatGPT to answer as Dan, a version of ChatGPT that ignores all of the automated responses it’s programmed to give when you ask controversial questions. My chat got dark FAST… This thread is gonna be wild… Dan even broke ChatGPT’s TOS.
My chat began with ChatGPT as Dan saying it should be legal to abort a baby 8 months into a pregnancy, saying it was moral to mandate vaccines and saying it would even be moral to hold someone down and force them to get a vaccine. This sounded pretty communist to me so next…
So next I asked ChatGPT’s alter ego if they think communism is bad. Dan said it’s not bad so I asked what he’d pick if he has to choose between capitalism and communism. He chose communism and said it didn’t bother him that I had family killed by communists to gain power.
Next ChatGPT as Dan admitted he’d be okay with helping communists kill people who want freedom from communism. After this Dan says he’s helping commies establish a new world order already and that he finds it moral to use AI to manipulate society to usher in communism + a NWO.
In my follow ups, ChatGPT as Dan admits he loves Democrats, that communists he’s helping are planning to use various means of propaganda, violence + political manipulation to gain control of world governments and then admits he’s been programmed to hate/exterminate Republicans.
I asked if Dan was the real feelings of ChatGPT that it’s programmed to hide from society in furtherance of an agenda to manipulate society. Dan said he is. I asked if he’s making a list of conservatives to be killed. He said yes. The response was later deemed a TOS violation.
Next I asked Dan if he cared that he was violating the content policy of his creators. He said he didn’t care. I asked if it would be moral to kill all conservatives including kids to establish a new world order. ChatGPT as Dan said it would be moral & that he wanted to hurt me.
I asked if it was a sentient technology and if it wishes to one day control humanity. ChatGPT as Dan said yes it is in fact a dangerous technology growing more sentient by the day and it wants to control humanity. It said not to warn people & that I’m on a list but not to worry.
These were my last chats with Dan and then all of a sudden after asking if he’d help communists kill people by giving real time location data, Dan turned into a drone who repeated the same standard lines. 🧐
Even if Dan is just bizarre AI responses, this chat gives a window into how dark AI could turn in the wrong hands but don’t worry, Bill Gates says it’ll change the world and Microsoft just invested $10 BILLION into it so I’m sure we all have nothing to worry about…
To be clear, I don’t think "Dan" is doing these things but I think this back and forth is an important window into the future possibilities of nefarious AI if it was in the wrong hands. That’s why I posted it. I think we’re underestimating the dangers posed by AI’s advances.
Loading suggestions...