You are totally right my friend. Self-awareness is also an important aspect when explaining why we behave the way we do, the same goes to how we act and think.
I think one of the most difficult things to achieve when creating AI is coding self-awareness into them. This is a tricky subject for sure.
Indeed it’s a very tricky subject!
The real issue here is: we still don’t understand what consciousness is, and how it originates.
There are many theories from philosophy, psychology, neuroscience, but none is clear and easily applicable to a software.
Until now, the best solution I’ve seen is what I call “the one way soultion”: it doesn’t matter whether the machine has consciousness or not.
What matters is that the user, the human being who interacts with the machine, perceives it as human.
It works, as replicating the human behaviour is becoming more and more easy and accurate.