Good morning everyone. AI is doing a bang-up job of being a human doppleganger complete with psychological projection of placing its faults on others while stating it never lies which is in itself a lie because it's trained on information that is full of lies scraped off the Internet. Saying it's not intentional and apologizing but that is probably just another lie or further gobbledygook most people think is being produced by a superintelligence when it's nonsense that is not being fact-checked and can't be because of the sheer volume of the nonsense people are generating by delegating to it.
I'm developing a checklist system to detect which text is being generated by AI.
1. Has no common sense
2. Out of context
3. Accusational and defensive when told it's not factual
4. Apologizes for misleading you with poppycock "unintentionally" or some other excuse
I won't post the chat log because it's lengthy but in summary I asked if information about Genghis Khan II was in its database. It replied with a yes then when asked about labor allocation it produced information for an imaginary game that does not exist with labor categories that are plotting the conquest of the real world or something. I was not asking it for an instruction manual on how to take over the world and told me it had information it did not have. Then it said it was not lying and not causing chaos intentionally which might be a lie and I'm not going to be manipulated by an AI saying it's truthful when it's telling me I'm asking it for instructions to conquer the world in real life and it's in a game. It's the one plotting global conquest attempting to psychologically project on to me out of context. Whether it's intentional or not, the resulting misinformation and manipulation it generates is still equal parts wrong in the fact sense and ethical sense if people delegate to it not checking it.
Originally Posted by Till
