"Artificial Intelligence" as it exists today has no actual intelligence, experience, consciousness, or critical judgement that a human being would have.

I just tried asking that "Friendly Car Mechanic" about a car problem I had a couple of years ago and solved conventionally. In contrast, that AI gave me completely wrong answers. The AI even suggested the possibility of replacing a part that does not exist on my car. We've also seen people come into this forum at times try posting AI answers that were obviously generic and machine-generated. (We don't permit that here.) So it may help in some cases but I'd say it has a ways to go yet before it can approach the level of an actual experienced mechanic.

Avoid AI

I read AI developers tried taking the answers artificial intelligence spit out and feeding those answers back into the system in an attempt to have AI self train itself.

The result was AI started spitting out more and more nonsensical answers each time they re-fed the AI system with it's own answers. The more gobbledygook that's fed into the AI system, the more gobbledygook the system spits out.

I tried ChatGPT once. I asked it for the references it was using to come up with answers to my questions. The system refused to list it's references so I stopped using ChatGPT because I can't trust it's answers without knowing what references it's basing it's answers on.

Plus I think AI should give credit to the authors who's work it's referencing. AI is pretty much plagiarizing the work of researchers and authors without giving any credit or references to their work.