Artificial Intelligence doesn't mean sentient life. It is generally regarded as intelligence that would be created initially by man [woman]. Once a machine was advanced enough it would start improving on it's own design [name the movie anyone?? lol].
The point when a machine starts to show intelligence is when it can learn from experience and modify it's behaviour based on the experience. Some early successes have been observed with spider like robots that have a limited "brain" but their program allows them to learn from their motions about the world around them and learning as they bump into things. The interesting thing is that the program was designed for them to not bump into things and to learn and adapt. Once they have discovered that there is an obstruction it is stored as a sort of living memory so to speak and they recall this and avoid it without the obstruction having been pre-programmed into them. This is an early development [and promising to the scientists] into AI although we won't be seeing a positronic brain for quite a few decades yet. When you start seeing AI built into cars, buildings, public transport it will be more in the vein of safety features -- but don't all Sci-Fi movies start like that??? lol
Actual self awareness is a totally different subject. When a machine can create, one might think that it has reached a level of sentience. Forget Turing and his "child machine" theory I don't think we will ever see that in any lifetime.
LF