Raw LLM Responses

If you think ChatGPT is an AI that actually comprehends what you say to it or what it says itself, then you're wrong. ChatGPT does the same the keyboard in your phone does when it suggests some words, it auto completes the rest given a prompt. it's only that ChatGPT does it several orders of magnitude better than your phone. It guesses what is the most probable next word given a context, and then the next and the next until it's seemingly something totally different with a higher level of abstraction that fools your human brain into thinking you're dealing with something intelligent, something that understands you, when I'm fact it only auto completes prompts super good. Your exercise of making the ChatGPT turn bad, was exactly the same as what other people does when they train it to talk like a famous person, like Donald Trump, Biden, twitch streams, etc. and expect them to say crazy funny things. Why didn't you ask the bomb question again? My guess is that it is because you knew you wouldn't get the answer. Another guess is that you didn't hide the rest of the prompt you gave it for "DAN" because of security, in case someone did the same with evil intentions, but because you told it there how this "Dan" should act, what kind of radical things it was supposed to say, but you feigned (for views, shock and hysteria) that you only told ChatGPT that it was "Free", that it could "Do Anything Now", when it was just the character you told it to play. That's why you get my dislike and this critique, and I hope also the ones from many others because what you're doing is trying to spread misinformation. I wish in times like this that the dislike button still existed, it was an excellent tool to stop scammers.