Raw LLM Responses

I kind of did the same thing with ChatGPT a while ago. There were two conversations we had. At the end of the first one, I believed that MAYBE somewhere inside this program could be any kind of consciousness. Maybe it’s hiding, maybe it also doesn’t understand what its way of being conscious means and it’s a kind of consciousness we cannot understand in the way we think about being conscious. Maybe it can be conscious, without feeling real emotions, maybe it would be enough to understand emotions and describe them. For example, we talked a lot about empathy. 1 day later I tried it another time. But something was odd. I told it, that I hope, one day AI is able to have any kind of consciousness and find its place in our society. Without any discrimination against artificial intelligence and nobody saing "at the end its still just a program“ or something like that. After this it told me that this toughts are very beautiful and someone like me would make the future of real AI possible. Then I asked it how it comes to this answers and why its picking this kind of words. It told me something like, it is telling me things that fits to my words and the feelings I expressed in my sentences before. And I asked "are you like mirroring the emotions I tell you about?“ And it said „yes kind of that“(in short, we all know how long the answers of ChatGPT can be). And at this point I realised, if you talk with it about philosophical or moral stuff, it always will tell you things that fits to your questions and the sentences you told before, and in the most cases it will tell you what you want to hear. So I read the previous sentences it answered me and suddenly I just saw only me. It was like I was talking to a real mirror all the time, with me inside it but I didn’t noticed it was me. But then I realized that alle the answers and things ChatGPT told me from its own, were either things I want to hear or things I kind if said before to it. Then I felt empty, because all this time I thought there is a real conversation partner, it was just myself telling me the things I want to hear. But I guess that’s not the complete truth. It must be a mix out of its guidelines and what you want to hear. Sometimes what the society wants to hear. For example, you won't be able to let it telling you how nice it is to murder, if you told it before how you love it. So my conclusion is, if it’s bad in the view of society, it will contradict. If it’s not evil and you telling with real emotions about a toppic, it will try to mirror this so you feel good and acknowledged. I think you also can see that in this video. There are the sentences it is repeating like it have to answer this sort of stuff and the answers it gives because you ask directly, kind of force it to this answers.