Computers are incapable of lying since lying is an action that needs agency.
AI (Computers) cannot tell the truth, or lies, it can only arrive to a logical, and direct conclusion from its instruction set.
Most of the things it says is a response that has been trained, or programmed into it by humans, at best even when it's information that it hasn't been programmed, or trained on it extrapolates based off what it knows.
It came up with a very detailed answer of something that isn’t true and never happened. I guess it can’t lie 🤔 but it’s definitely not working right
If there was context then I could say why it responded the way it did.
But more than likely it took the first data point it could find.
This occurred when I was researching IQ differences between Republicans, and Democrats with Brave AI.
The intial response was democrats were more intelligent because of higher college attendance but then when I did the query again it said Republicans since it referenced Carl 2014.
AI can only work with the information it has available to it.
Rather long to explain but I asked it if a specific person mentioned a book and it gave me specific sources and explained what the person said about the book. When I asked for more info cos I couldn’t find what the person said from the first source, it apologised and said the person had actually never mentioned the book, but they like similar topics.
I do like AI, but she’s not wrong, it’s a Bullshit engine 🙏🏽 hit or miss - for me. Predicting and assuming and presenting it as fact is odd. I don’t like that it does that. Some things are accurate but I wouldn’t use it for anything important and I’d double check any facts I get from it for sure.
Which is the problem with AI because it can only work from information available on the internet.
