Well I guess this is where the confusion sets in...The mainstream notion of AI is, that if your no longer able to tell, that your chatting with a program and not a human. Than that is equal to true AI.
It is not!
This is really VI... Which can be extremely complex, but it will never be true AI.
It is worth noting here, that as far as I know, no one has the slightest idea how to develop an Quantum AI...
We haven't even been able to construct convincing neuronet models using conventional computer mechanics yet.
A QC may make this neuronet integration easier with time, but no one knows how.
And let's just get the 1st iteration of the thing first, before the imagination starts to run wild
Edit:
What it currently looks like a QC would be excellent at, would be f.eks spacenavigation. Very very very complex calculations, can be answered with very good probability, near instantly... You do them over and over, to keep on course... Could be very efficient.
In research of all fields the currently envisioned QC would be invaluable! You could simply not compete without it...